US20230396709A1 - System and method for mobile device active callback prioritization with predictive outcome scoring - Google Patents

Abstract

A system and methods for mobile device active callback prioritization with predictive outcome scoring, utilizing an enhanced callback prioritization engine operating on a user's mobile device for integration through the operating system and software applications operating on the device, wherein the enhanced callback prioritization engine receives intercepted data or voice messages sent to the mobile device, retrieves and aggregates data related to the assigned messages, inputs the assigned data message and aggregate data into prioritization model wherein the prioritization model may analyze the input data, the results of the analysis may be used to compute a priority score for the assigned data message. An outcome model may be present and configured to produce outcome scores for a callback recipient based on the context of the data or voice message. System generates a callback list from the computed prioritization score and outcome score.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS
• Priority is claimed in the application data sheet to the following patents or patent applications, each of which is expressly incorporated herein by reference in its entirety:
• • Ser. No. 18/152,705
  • Ser. No. 17/750,918
  • Ser. No. 17/572,405
  • Ser. No. 17/389,837
  • Ser. No. 16/985,093
  • Ser. No. 16/583,967
  • Ser. No. 62/828,133
  • Ser. No. 16/542,577
  • Ser. No. 62/820,190
  • Ser. No. 16/523,501
  • Ser. No. 15/411,424
BACKGROUND OF THE INVENTIONField of the Art
• The disclosure relates to the field of mobile device technology, specifically to the field of integrating cloud-based callback management systems with onboard software on a mobile device.
Discussion of the State of the Art
• Typically, if a client calls a business, voice prompt menu choices enable the calling client to identify an issue for which the client requires service and the client is then queued for a service agent capable of handling the identified issue. As such, it is expected that clients who identify the purpose of their call as a “billing issue” will be queued for, and connected to, a service representative with the ability to handle billing issues. Similarly, it is expected that clients who identify the purpose of their call as a “customer service issue” will be queued for, and connected to, a service representative with the ability to handle customer service issues.
• There are problems with existing communications systems, such as contact centers, including the following two problems. First, the voice prompt menus that are used to channel callers to the queue for the appropriate group of service agents are frustrating to clients. It takes significant time to navigate the layered menus of voice prompts.
• Second, waiting on-hold while a connection, be it a phone call, web chat, video conference, or other interaction type, is maintained in queue for connection to a service agent is also frustrating to clients.
• In an effort to reduce customer exacerbation caused by having to maintain a connection while on-hold in queue, secondary queue systems have been developed. A typical secondary queue system obtains a telephone number at which the calling client can be reached when a service representative is available (i.e., a call back number). The client disconnects, and then, at the proper time, a call back system establishes a connection to the client utilizing the call back number and couples the client to an available representative without waiting on-hold in queue. One exemplary system is disclosed in U.S. Pat. No. 6,563,921 to Williams et al. which is commonly assigned with the present application.
• While such a system may make the experience of waiting for a connection to a service representative less frustrating, it does not address the inconvenience of having to navigate a slow and complicated voice prompt menu to enter the queue.
• Additionally, as the gig economy compounds and people tend to get busier in the professional workplace, a solution is needed on their personal device to help prioritize returning communications with personal contacts, business relations, or solicitors.
• What is needed is a system and various methods for providing integration of a cloud callback platform with mobile device software so that callback functionality and prioritization becomes a transparent and consistent feature across interaction modes through the mobile device ecosystem.
SUMMARY OF THE INVENTION
• Accordingly, the inventor has conceived and reduced to practice, a system and methods for mobile device active callback prioritization with predictive outcome scoring, utilizing an enhanced callback prioritization engine operating on a user's mobile device for integration through the operating system and software applications operating on the device, wherein the enhanced callback prioritization engine receives intercepted data or voice messages sent to the mobile device, retrieves and aggregates data related to the assigned messages, inputs the assigned data message and aggregate data into prioritization model wherein the prioritization model may analyze the input data, the results of the analysis may be used to compute a priority score for the assigned data message. An outcome model may be present and configured to produce outcome scores for a callback recipient based on the context of the data or voice message. System generates a callback list from the computed prioritization score and outcome score.
• According to a first preferred embodiment, a mobile device with active callback prioritization is disclosed, comprising: a processor, a memory, and a plurality of programming instructions stored in the memory and operable on the processor; a callback integration engine comprising a subset of the plurality of programming instructions that, when operating on the processor, cause the processor to: receive a data or voice message, the data or voice message comprising at least one characteristic; produce a callback object in memory comprising information associated with the data or voice message received; and send the callback object to an enhanced callback prioritization engine; and the enhanced callback prioritization engine comprising a subset of the plurality of programming instructions that, when operating on the processor, cause the processor to: receive the callback object from the callback integration engine; retrieve and aggregate application data related to the data or voice message; retrieve callback recipient data for a plurality of callback recipients; use the callback object and the aggregated application data as inputs into a prioritization model, wherein the prioritization model analyzes the callback object's information and the aggregated application data to determine the context and urgency associated with the data or voice message; for each callback object, compute a priority score based at least upon the results of the analysis; use the determined context and urgency and the retrieved callback recipient data as inputs into one or more machine learning algorithms, wherein the machine learning algorithms analyze the context and urgency and the retrieved callback recipient data to determine an interaction outcome associated with the data or voice message; for each callback recipient, compute an outcome score based at least upon the results of the analysis; assign a callback recipient to the callback object based at least upon the outcome score; and use the computed priority score, the callback object data, and the data or voice message to generate a callback list.
• According to a second preferred embodiment, a method for active callback prioritization is disclosed, comprising the steps of: receiving a data or voice message, the data or voice message comprising at least one characteristic; producing a callback object in memory comprising information associated with the data or voice message received; sending the callback object to an enhanced callback prioritization engine; receiving the callback object from the callback integration engine; retrieving and aggregate application data related to the data or voice message; retrieving callback recipient data for a plurality of callback recipients; using the callback object and the aggregated application data as inputs into a prioritization model, wherein the prioritization model analyzes the callback object's information and the aggregated application data to determine the context and urgency associated with the data or voice message; for each callback object, computing a priority score based at least upon the results of the analysis; using the determined context and urgency and the retrieved callback recipient data as inputs into one or more machine learning algorithms, wherein the machine learning algorithms analyze the context and urgency and the retrieved callback recipient data to determine an interaction outcome associated with the data or voice message; for each callback recipient, computing an outcome score based at least upon the results of the analysis; assigning a callback recipient to the callback object based at least upon the outcome score; and using the computed priority score, the callback object data, and the data or voice message to generate a callback list.
• According to an aspect of an embodiment, the enhanced callback prioritization engine is further configured to: retrieve and aggregate requestor-specific data; use the aggregated requestor-specific data as an additional input into the one or more machine learning algorithms, wherein the machine learning algorithms analyze the context and urgency data, the retrieved callback recipient data, and the aggregated requestor-specific data to determine an requestor-specific outcome associated with the data or voice message; for each callback recipient, compute a requestor-specific outcome score based at least upon the results of the analysis; and assign a callback recipient to and the callback object based at least upon the requestor-specific outcome score.
• According to an aspect of an embodiment, the callback recipient data comprises agents skills, quality assurance scores, scheduling information, historical interactions, spoken languages, certifications, interaction outcomes, and location information.
• According to an aspect of an embodiment, the requestor-specific data comprises requestor preferences.
• According to an aspect of an embodiment, the requestor preferences comprise preferred agents or agent qualities.
• According to an aspect of an embodiment, the application data is retrieved using application programming interfaces.
• According to an aspect of an embodiment, the callback list comprises a smart reply message.
• According to an aspect of an embodiment, the callback integration engine receives a user confirmed callback list and executes the confirmed callback items on the list.
• According to an aspect of an embodiment, the application data comprises data from communication, social media, financial, gaming, and productivity applications.
• According to an aspect of an embodiment, the machine learning algorithms comprise natural language processing.
BRIEF DESCRIPTION OF THE DRAWING FIGURES
• The accompanying drawings illustrate several aspects and, together with the description, serve to explain the principles of the invention according to the aspects. It will be appreciated by one skilled in the art that the particular arrangements illustrated in the drawings are merely exemplary, and are not to be considered as limiting of the scope of the invention or the claims herein in any way.
• FIG.1 is a block diagram illustrating an exemplary system architecture for a mobile device connected to a cloud callback platform via a public switch telephone network and the Internet, according to an embodiment.
• FIG.2 is a block diagram illustrating an exemplary mobile device with a callback integration engine and callback prioritization engine operating at the operating system level, according to an embodiment.
• FIG.3 is a block diagram illustrating an exemplary mobile device with a callback integration engine and callback prioritization engine operating at the device firmware level, according to an embodiment.
• FIG.4 is a block diagram illustrating an exemplary system architecture for a callback prioritization engine, according to an embodiment.
• FIG.5 is a user interface diagram illustrating an exemplary callback list displayed on a mobile device with integrated callback features, according to an embodiment.
• FIG.6 is a user interface diagram illustrating an exemplary incoming call on a mobile device with integrated callback features, according to an embodiment.
• FIG.7 is a user interface diagram illustrating an exemplary email application operating on a mobile device with integrated callback features, according to an embodiment.
• FIG.8 is a method diagram illustrating an exemplary incoming call flow, according to an embodiment.
• FIG.9 is a method diagram illustrating an exemplary callback workflow once a user selects a callback for an incoming call, according to an embodiment.
• FIG.10 is a method diagram illustrating an exemplary callback prioritization workflow once a data or voice message is received, according to an embodiment.
• FIG.11 is a block diagram illustrating an exemplary system architecture for an exemplary mobile device embodiment with an enhanced callback prioritization engine operating at the operating system level and connected to a cloud callback platform, according to an embodiment.
• FIG.12 is a block diagram illustrating an exemplary system architecture for an enhanced callback prioritization engine, according to an embodiment.
• FIG.13 is a method diagram illustrating several app integrations for enhancing callback prioritization.
• FIG.14 is message flow diagram illustrating an exemplary exchange of data messages between components of a cloud callback platform with active callback prioritization, according to an embodiment
• FIG.15 is message flow diagram illustrating an exemplary exchange of data messages between components of a cloud callback platform with active callback prioritization and requestor-specific outcome scoring, according to an embodiment.
• FIG.16 is a flow diagram illustrating an exemplary method for training and testing a machine learning model, according to an embodiment.
• FIG.17 is a block diagram illustrating an exemplary hardware architecture of a computing device.
• FIG.18 is a block diagram illustrating an exemplary logical architecture for a client device.
• FIG.19 is a block diagram showing an exemplary architectural arrangement of clients, servers, and external services.
• FIG.20 is another block diagram illustrating an exemplary hardware architecture of a computing device.
DETAILED DESCRIPTION OF THE DRAWING FIGURES
• The inventor has conceived, and reduced to practice, a system and methods for mobile device active callback prioritization with predictive outcome scoring, utilizing an enhanced callback prioritization engine operating on a user's mobile device for integration through the operating system and software applications operating on the device, wherein the enhanced callback prioritization engine receives intercepted data or voice messages sent to the mobile device, retrieves and aggregates data related to the assigned messages, inputs the assigned data message and aggregate data into prioritization model wherein the prioritization model may analyze the input data, the results of the analysis may be used to compute a priority score for the assigned data message. An outcome model may be present and configured to produce outcome scores for a callback recipient based on the context of the data or voice message. System generates a callback list from the computed prioritization score and outcome score.
• One or more different aspects may be described in the present application. Further, for one or more of the aspects described herein, numerous alternative arrangements may be described; it should be appreciated that these are presented for illustrative purposes only and are not limiting of the aspects contained herein or the claims presented herein in any way. One or more of the arrangements may be widely applicable to numerous aspects, as may be readily apparent from the disclosure. In general, arrangements are described in sufficient detail to enable those skilled in the art to practice one or more of the aspects, and it should be appreciated that other arrangements may be utilized and that structural, logical, software, electrical and other changes may be made without departing from the scope of the particular aspects. Particular features of one or more of the aspects described herein may be described with reference to one or more particular aspects or figures that form a part of the present disclosure, and in which are shown, by way of illustration, specific arrangements of one or more of the aspects. It should be appreciated, however, that such features are not limited to usage in the one or more particular aspects or figures with reference to which they are described. The present disclosure is neither a literal description of all arrangements of one or more of the aspects nor a listing of features of one or more of the aspects that must be present in all arrangements.
• Headings of sections provided in this patent application and the title of this patent application are for convenience only and are not to be taken as limiting the disclosure in any way.
• Devices that are in communication with each other need not be in continuous communication with each other, unless expressly specified otherwise. In addition, devices that are in communication with each other may communicate directly or indirectly through one or more communication means or intermediaries, logical or physical.
• A description of an aspect with several components in communication with each other does not imply that all such components are required. To the contrary, a variety of optional components may be described to illustrate a wide variety of possible aspects and in order to more fully illustrate one or more aspects. Similarly, although process steps, method steps, algorithms or the like may be described in a sequential order, such processes, methods and algorithms may generally be configured to work in alternate orders, unless specifically stated to the contrary. In other words, any sequence or order of steps that may be described in this patent application does not, in and of itself, indicate a requirement that the steps be performed in that order. The steps of described processes may be performed in any order practical. Further, some steps may be performed simultaneously despite being described or implied as occurring non-simultaneously (e.g., because one step is described after the other step). Moreover, the illustration of a process by its depiction in a drawing does not imply that the illustrated process is exclusive of other variations and modifications thereto, does not imply that the illustrated process or any of its steps are necessary to one or more of the aspects, and does not imply that the illustrated process is preferred. Also, steps are generally described once per aspect, but this does not mean they must occur once, or that they may only occur once each time a process, method, or algorithm is carried out or executed. Some steps may be omitted in some aspects or some occurrences, or some steps may be executed more than once in a given aspect or occurrence.
• When a single device or article is described herein, it will be readily apparent that more than one device or article may be used in place of a single device or article. Similarly, where more than one device or article is described herein, it will be readily apparent that a single device or article may be used in place of the more than one device or article.
• The functionality or the features of a device may be alternatively embodied by one or more other devices that are not explicitly described as having such functionality or features. Thus, other aspects need not include the device itself.
• Techniques and mechanisms described or referenced herein will sometimes be described in singular form for clarity. However, it should be appreciated that particular aspects may include multiple iterations of a technique or multiple instantiations of a mechanism unless noted otherwise. Process descriptions or blocks in figures should be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps in the process. Alternate implementations are included within the scope of various aspects in which, for example, functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those having ordinary skill in the art.
Definitions
• “Callback” as used herein refers to an instance of an individual being contacted after their initial contact was unsuccessful. For instance, if a first user calls a second user on a telephone, but the second user does not receive their call for one of numerous reasons including turning off their phone or simply not picking up, the second user may then place a callback to the first user once they realize they missed their call. This callback concept applies equally to many forms of interaction that need not be restricted to telephone calls, for example including (but not limited to) voice calls over a telephone line, video calls over a network connection, or live text-based chat such as web chat or short message service (SMS) texting. While a callback (and various associated components, methods, and operations taught herein) may also be used with an email communication despite the inherently asynchronous nature of email (participants may read and reply to emails at any time, and need not be interacting at the same time or while other participants are online or available), the preferred usage as taught herein refers to synchronous communication (that is, communication where participants are interacting at the same time, as with a phone call or chat conversation).
• “Callback object” as used herein means a data object representing callback data, such as the identities and call information for a first and second user, the parameters for a callback including what time it shall be performed, and any other relevant data for a callback to be completed based on the data held by the callback object.
Conceptual Architecture
• FIG.1 is a block diagram illustrating anexemplary system architecture100 for amobile device101 connected to acloud callback platform110 via a publicswitch telephone network103 and theInternet102, according to an embodiment. According to the embodiment,callback cloud110 may receive requests via a plurality of communications networks such as a public switched telephone network (PSTN)103 or theInternet102. These requests may comprise a variety of communication and interaction types, for example including (but not limited to) voice calls over a telephone line, video calls over a network connection, or live text-based chat such as web chat or short message service (SMS) texting viaPSTN103. Such communications networks may be connected to a plurality ofmobile devices101 such as a user's smartphone or similar mobile device, according to the particular architecture of communication network involved.Mobile devices101 may be connected to respective communications networks via a variety of means, which may include telephone dialers, VOIP telecommunications services, web browser applications, SMS text messaging services, or other telephony or data communications services. It will be appreciated by one having ordinary skill in the art that such means of communication are exemplary, and many alternative means are possible and becoming possible in the art, any of which may be utilized as an element ofsystem100 according to the invention.
• When needed (for example, when a user manually requests a callback or when configured settings determine that a callback is needed), a user'smobile device101 creates a session with acloud callback platform110 with a profile manager111, resulting in a callback being required. Profile manager111 receives initial requests to connect tocallback cloud110, and forwards relevant user profile information to acallback manager113, which may further request environmental context data from anenvironment analyzer112. Environmental context data may include (for example, and not limited to) recorded information about when a user may be suspected to be driving or commuting from work (as may be determined from device information, such as whether a “do not disturb while driving” feature has been enabled, or if themobile device101 is currently connected to a car audio system), or if the user may be busy for another reason, such as if they are running or working out (as may be determined from device sensor data), for example, and may be parsed from online profiles or online textual data, using anenvironment analyzer112.
• Acallback manager113 centrally manages all callback data, creating a callback object which may be used to manage the data for a particular callback, and communicates with aninteraction manager114 which handles requests to make calls and bridge calls, which go out to amedia server115 which actually makes the calls as requested. In this way, themedia server115 may be altered in the manner in which it makes and bridges calls when directed, but thecallback manager113 does not need to adjust itself, due to going through an intermediary component, theinteraction manager114, as an interface between the two. Amedia server115, when directed, may place calls and send messages, emails, or connect voice over IP (“VoIP”) calls and video calls, to users over aPSTN103 or theInternet102.Callback manager113 may work with a user's profile as managed by a profile manager111, with environmental context from anenvironment analyzer112 as well as (if provided) EWT information for any callback recipients (for example, contact center agents with the appropriate skills to address the callback requestor's needs, or online tech support agents to respond to chat requests), to determine an appropriate callback time for the two users (a callback requestor and a callback recipient), interfacing with aninteraction manager114 to physically place and bridge the calls with amedia server115.
• Thecloud callback platform110 may further include an application programming interface (API)manager116 and acallback prioritization engine117, according to an embodiment. TheAPI manager116 may be used to expose or connect with software related to a callback requester and a callback recipient. For example,API manager116 may be used to connect with email, messaging, telephone, or other voice communication applications present on a mobile device in order to retrieve text-based data related to emails, text messages such as SMS and from other messaging applications, or voice data such as voicemails. Voice data such as, for example, a voice mail may be transcribed into a text-based data message by thecloud callback platform110. The retrieved text-based data messages may be sent to aprioritization engine117 which may identify one or more characteristics of the text-based data messages to route the data messages to an appropriate lobby. Theprioritization engine117 may use the environment analyzer to parse or analyze the retrieved text-based data messages. Analysis of the text-based data messages may be used to determine or infer a variety of metrics pertaining to a callback requester and at least one callback recipient, such as (but not limited to) relationships, intent, subject matter, and sentiment. Based on the analysis of the text-based data messages, thecallback prioritization engine117 may produce priority score for each text-based data message which can be used to order lobbies (callback groups) in order of priority.
• FIG.2 is a block diagram illustrating an exemplarymobile device101 with acallback integration engine201 andcallback prioritization engine202 operating at the operating system level, according to an embodiment. According to the embodiment, amobile device101 comprises several hardware and software components operating at various levels to provide various functions. At the mostbasic hardware level210, the physical hardware of themobile device101 may comprise a processor andmemory211 that provide standard computing device functionality as described in detail below, with reference toFIGS.14-17 (grouped together here for the sake of clarity, it should be appreciated that processor/memory may or may not be part of the same hardware component, such as a system-on-a-chip), abaseband processor212 that manages radio-based communication functions such as cellular or Wi-Fi connectivity, as well as any of a number ofhardware sensors213 such as (for example, including but not limited to) an accelerometer to detect or measure device movement, gyroscope for detecting or measuring device orientation, barometer for measuring ambient environment conditions, or global positioning system (GPS) receiver for geolocating the device.
• An operating system (OS)220 comprises the main software operating onmobile device101 and providing various software-based functions such as software applications and communications.Various software applications221 operating within (for example in an application layer not shown here but as is commonly used in various computing devices according to the standard OSI process model of computing systems) on themobile device101 may expose and provide access to, or interaction with, various hardware and sensor features such as to enable a user to view or calibrate sensor readings.
• According to the embodiment, acallback integration engine201 may comprise local (that is, operating on the mobile device) functionality similar to a cloud-basedcallback manager113 described above, and may operate as an application or feature at theoperating system level220, running at a similar privilege level and having similar access to hardware and software functions asother applications221 operating onmobile device101. This enables easy installation or removal ofcallback integration engine201, as it may be readily distributed via similar methods as any other software application (for example, via an application store or similar download portal). However, this may restrict the capabilities ofcallback integration engine201, as it will have only indirect access to hardware features (as it will only be able to access whatever features are exposed by the operating system) and in many cases may be “sandboxed”, and prevented from directly accessing other software or information on the mobile device201 (for example, in a sandboxed software environment such as APPLE IOS™ or similar).
• According to the embodiment, acallback prioritization engine202 may comprise local (that is, operating on the mobile device) functionality similar to a cloud-basedprioritization engine117 described above, and may operate as an application or feature at theoperating system level220, running at a similar privilege level and having similar access to hardware and software functions asother applications221 operating onmobile device101. Thecallback prioritization engine202 may be included within the same application as thecallback integration engine201 or it may be packaged as its own application with built in functionality to communicate with and share data with a separatecallback integration engine201. This enables easy installation or removal ofcallback prioritization engine202, as it may be readily distributed via similar methods as any other software application (for example, via an application store or similar download portal). However, this may restrict the capabilities ofcallback prioritization engine202, as it will have only indirect access to hardware features (as it will only be able to access whatever features are exposed by the operating system) and in many cases may be “sandboxed”, and prevented from directly accessing other software or information on the mobile device201 (for example, in a sandboxed software environment such as APPLE IOS™ or similar).
• Callback prioritization engine202 may receive from the callback integration engine201 a callback object associated with intercepted data or voice messages.Callback prioritization engine202 may intercept (e.g., receive, hook, access, or otherwise obtain) voice messages (i.e., voicemail) and text-based data messages such as (but not limited to) short message service (SMS), emails, and group chats sent to themobile device101 and route them to an appropriate lobby (e.g., group) based upon at least one characteristic of the message. In an embodiment, the mobile device may transcribe voice messages into text-based data message for further processing. Each text-based data message and voice message (or its transcription) may have one or more characteristics, such as indications of source, destination, data message subject matter, end user identifiers or account identifiers. Once a data message has been assigned to a lobby other data associated with the data message may be aggregated and linked to the callback object. For example, a first email sent to amobile device101 may be intercepted and assigned to a lobby, and then thecallback prioritization engine202 can identify and retrieve other emails, text messages, or group chats associated with the first email in order to provide more context for the first email message. In this example, other communications exchanged between the first email recipient and the first email sender (such as an email thread, text messages, or voicemails) may be aggregated and added to the callback object associated with the first email. The aggregated data may then be fed into one or more machine learning algorithms that analyze the aggregated data and can generate for each data message in the lobby a priority score. The priority score may be used to create a hierarchy or priority list of callback objects such that each callback may be scheduled and executed in order of highest priority to lowest priority.
• FIG.3 is a block diagram illustrating an exemplarymobile device101 with acallback integration engine201 andcallback prioritization engine202 operating at the device firmware level, according to an embodiment. According to the embodiment, afirmware level320 encompasses low-level program code that operates “below” theoperating system220, operating directly on hardware components of mobile101, and comprises such features as a baseband controller321 that controls features ofbaseband processor212 and the operation of which is fully transparent to the user (that is, a user ofmobile device101 does not see or interact with firmware, and many users may not even be aware of its existence or capabilities).
• According to the embodiment, acallback integration engine201 and acallback prioritization engine202 may operate as a firmware component that directly interfaces with hardware features ofmobile device101, enabling direct control as needed while exposing desired functionality to theoperating system220 in a transparent manner (that is, the operating system may only know that a feature is available, and may not be able to determine that the feature is provided by thecallback integration engine201 orcallback prioritization engine202 rather than, for example, the baseband controller321). This enables more direct control over device functions, such as call and data message routing and hardware information such as sensor readings (as may be used in callback workflow processing), and allows thecallback integration engine201 to intercept incoming calls and radio information (e.g., data messages) before it reaches the operating system, so data may be manipulated and additional features may be integrated prior to presentation to the OS. This also prevents accidental removal ofcallback integration engine201 andcallback prioritization engine202, as a user may have no control over device firmware and anything operating at or below theOS level220 would inherently have insufficient access to modify firmware. This further enablescallback integration engine201 andcallback prioritization engine202 to access hardware features that may be restricted or completely prevented when operating within theOS layer220, for example hardware sensor or geolocation information that may be incorporated into a callback workflow.
• FIG.4 is a block diagram illustrating anexemplary system architecture400 for acallback prioritization engine202, according to an embodiment. According to the embodiment, acallback prioritization engine202 may receive a data message with a characteristic. The data message can part of sessions that include asynchronous text, voice message, or data message (e.g., not a live voice conversation such as a telephone call) communications between and end user and a computing device (e.g., mobile device, contact center agent [e.g., a person, bot, or combination thereof], etc.). In some examples,callback prioritization engine202 can receive voice messages that have been send as audio messages (voicemail) or converted into text messages. The sessions can occur continuously in a conversational manner, or can be intermittent, e.g., lasting over hours, days or weeks with dormant periods between communications. Thecallback prioritization engine202 may receive data messages and can identify characteristics of those data messages.
• Based upon the characteristics of the data messages, thelobby assignment mechanism401 can assign the data message to a default lobby construct (group) or can bypass the default lobby construct and instead assign the data message to the destination lobby construct. For example. Thelobby assignment mechanism401 can determine that the characteristics of a data message do not match or correlate to characteristics of other data messages that have been assigned to one or more destination constructs. When thelobby assignment mechanism401 does identify a match or correlation between a (new incoming) data message and a data message assigned to a particular destination lobby thelobby assignment mechanism401 can assign the new incoming data message the same destination lobby. The lobby constructs hold collections of data messages corresponding to communications or sessions between two or more computing devices (e.g., an end user mobile device and a contact center agent computing device or coworker computing device). The plurality of lobby constructs can be identical instances or types of structures, differing only in the collections of data messages they hold and the teams to which they are assigned. For example, there can be exactly one lobby construct assigned per defined team of contact center agent computing devices to attend to defined issues, such as customer complaints or service issues. Similarly, there can be exactly one lobby construct assigned per defined teams (e.g., marketing, sales, customer retention, managerial staff, etc.) within a business structure or a lobby construct may be assigned to a single person or entity. The collections of data messages for this team may be held by a single lobby construct.
• Once a data message has been assigned to a lobby construct thecallback prioritization engine202 may identify and aggregate other data related to the data message. For example, if an assigned data message is part of an ongoing session (such as an email thread, text or group messaging conversation) then adata aggregator402 may retrieve the other messages contained within the session and append or link the other messages to the assigned data message. Additionally, if the assigned data message is a stand-alone message (i.e., sent with no connection to a prior session) then thecallback prioritization engine202 may scan for and identify other communications that may be related to the assigned data message. Other communications may be identified as related based upon one or more characteristics of the assigned data message. For example, an end user may receive an email asking for a response of some sort from a coworker which may be related to previous communications exchanged between the end user and the coworker, but not contained within the same thread as the previous communications. In this example, thecallback prioritization engine202 may be able to identify the previous communications as related to the original email asking for a response and append or link them to the assigned data messages generated from the response email. As another example, an end usermobile device101 may receive a voicemail from an individual or entity (business) asking the end user to callback when available. The voicemail may be transcribed to a text data message and have one or more characteristics, such as the phone number or other identifier associated with the voicemail, and thedata aggregator402 can use the voicemail characteristics to identify an email (or other data message) that may be associated with the voicemail. For example, an employee of a company may send to a second employee a voicemail requesting a callback and further, the two employees may have been emailing or text messaging each other earlier in the day or week. These emails or text messages may be linked to the transcribed voicemail by thedata aggregator402. In this way thedata aggregator402 may identify and aggregate data correlating with a data message assigned to a lobby construct.
• The assigned data message and any linked and aggregated data may then be sent as input to a machine learning (ML)engine403 for processing and analysis. TheML engine403 may comprise one or more machine learning algorithms such as, for example, a natural language processing (NLP) neural network. The NLP may be used for text classification functions including, but not limited to, sentiment analysis, topic categorization, language recognition, and intent analysis. The NLP may be used to perform sentiment analysis on the assigned data message and any available aggregated data linked to the data message. Sentiment analysis may be used to determine if the content of a data message (and its aggregate data) is positive, negative, or neutral. This could be used to classify data messages to better respond to inquiries or requests. The NLP may be used for topic categorization to determine which category (or lobby construct) a data message belongs to, essentially this constitutes a robust, automatic method of tagging and organizing the content of a data message or its aggregate data. This may be used to automatically separate data messages within a lobby construct by topic. The NLP may be used to understand the intent of a data message or its aggregate data by determining the underlying intention of a given data message. Additionally, the NLP may be used to find and identify relationships between a caller and callee, or a message recipient and a message sender, or between and among a group of people (i.e., multiple people in an email thread or group text message). Using the identified relationships,callback prioritization engine202 orcallback integration engine201 may generate personalized smart replies based on the identified relationships and other information (intent, sentiment, topic) derived from the NLP. The relationships may be used to determine the priority of an assigned data message within a lobby construct.
• In an embodiment, theML engine403 may also comprise one or more audio machine learning algorithms which may be used for audio classification and speech recognition functions. Audio classification ML algorithms may be used when a voice messages (voicemail) is received bymobile device101 or bycallback prioritization engine202. The audio algorithm may be used to identify a speaker by taking a snippet of the voice message to create a unique identifier, like a voice fingerprint. Then the fingerprint may be input into a neural network that can match the unique identifier to a known or previously heard voice. This may allowcallback prioritization engine202 to identify callers who may be calling from a new device or using a new identifier that may not be known by thecallback integration engine201. Furthermore, an audio ML algorithm may be utilized for speech recognition functions such as voice to text technology (i.e., transcribing a voicemail into a text data message) or to provide voice assistant interaction (e.g., in a contact center environment whereby a customer may communicate with a bot or script voice assistant).
• There are various benefits of having aML engine403 operating locally on a mobile device compared to operating in a cloud environment which the mobile device may connect to. Operating theML engine403 on amobile device101 protects against low latency. As most ML algorithms may be operating in real-time, network speed may be too slow to transmit a live data feed from the device, to thecloud callback platform110, and then back to the device. Furthermore, running theML engine403 locally on amobile device101 removes the risk of low or no network connections halting the ML processes. In amobile device101 theML engine403 may still work when run on device, even if low or no connectivity. Operating theML engine403 locally on amobile device101 facilitates privacy over the mobile device user's data. ML needs data to build a useful model and to make useful inferences derived from the model, however, some data may be sensitive (e.g., personal identifying information, regulated data). Thecallback prioritization engine202 can use and secure data on the device, without having to transmit any sensitive data.
• TheML engine403 may forward the results of the analyses (e.g., inferences, intent, sentiment, topic, etc.) of the various ML algorithms to aprioritization module404. Theprioritization module404 may also receive, from thelobby assignment mechanism401, the assigned lobby construct(s) comprising at least one or more assigned data messages. Theprioritization module404 may then use the results of the analyses (e.g., inferences, intent, sentiment, topic, etc.) to generate a priority score for each data message assigned to a lobby construct. For example, if multiple assigned messages within a construct have been identified to belonging to the same topic, then the most recent data message pertaining to the shared topic may be given a higher priority score and thus the callback associated with that data message would be executed before the previously received data messages. For example, a cable company service agent may be assigned to a lobby construct which has been assigned twenty five data messages, and of those data messages twenty are about a loss of cable service. In this example, theML engine403 may be able to identify that each of those twenty data messages originated from the same area code and therefore there must be some service disruption affecting that area code. Theprioritization module404 can use this information to move the first of the twenty people up in priority so that the problem with their cable service may be addressed. If the cable service agent is able to resolve the issue for the first of twenty data messages relating to cable service loss, then thecallback prioritization engine202 may be able to send an automated message to the other of the customers associated with the remaining nineteen data messages providing information about the loss of cable service. For example, an automated message may be sent which says, “Is your call about loss of cable service? If so, we are aware and your service should be restored soon. Would you like to request a callback?”. If the receiver of the automated message replies with a callback request, then the data message may remain in the lobby construct, otherwise the data message will be removed from the lobby construct.
• Intent and sentiment analysis results may be used to determine priority as well. For example, the sentiment of one data message in a lobby construct may be determined to be positive, whereas the sentiment of another data message in same lobby construct may be urgent, or negative. A data message with an urgent or negative sentiment may be given a higher priority score than a data message with a positive sentiment in order to account for the urgency of the data message, or to quickly respond to the negative sentiment data message in order to change sentiment of the data message sender to neutral or positive. Additionally, the intent of a data message may be used to determine the priority of response. For example, a data message in a lobby construct may indicate that there was no intent for a callback or response, whereas another data message in the lobby construct may have an inferred intent for a callback. In this case, the data message with the inferred intent for a callback will be given a higher priority score than the data message with little to no inferred callback intent.
• Based upon the priority score of each data message in a lobby, theprioritization module404 may reorder the lobby in a hierarchal manner where the highest priority data message is at the top of the lobby and the other data messages are sorted in descending order of priority. Theprioritization module404 may also produce a callback list based upon the sorted lobby. The callback list may include the original data message, any aggregate data, and data contained within the callback object associated with each data message. For example, the callback list may display the original data message, a scheduled callback time, and the name of the callback recipient. The callback list may be sent to the mobile device101 (or any computing device) for display and interaction by the mobile device user.
• FIG.5 is a user interface diagram illustrating anexemplary callback list500 displayed on amobile device101 with integrated callback and prioritization features, according to an embodiment. According to the embodiment, a mobile device with integrated callback and prioritization features provided by acallback prioritization engine202 which may generate a callback list from received data messages or voice messages. For example, amobile device101 user may be unable to respond to a phone (or voice over internet protocol, “VoIP”) call, email, text, or other functionality. A user may be unable to respond for a plurality of reason such as, for example, the user may already be in a call, the user is in a meeting, the user is driving, or the user is otherwise disposed (e.g., working out, doing yard work, etc.). In these instances, thecallback integration engine201 may create a callback object for each received data message or phone call. The callback objects may be sent to acallback prioritization engine202 where the callback objects may be appended with additional data relating to the received data message or phone call. The data messages or associated callback objects may be assigned to a lobby construct (i.e., callback group). The data messages may then be analyzed via one or more ML algorithms to further provide, determine, generate, or infer context information that may be used to compute priority scores for each data message or callback object in a lobby construct. Acallback list500 may be generated from the lobby constructs, the data messages or callback objects contained with a lobby construct, and the computed priority scores to produce a hierarchal callback list based on priority. In this way, when amobile device101 user, who was otherwise unable to respond to the data messages and phone calls, is able to interact with the mobile device101 (s)he may be able to view an automatically generated callback list with tentative callback times scheduled.
• In an embodiment, the callback list may be displayed on amobile device101 allowing a mobile device user to view and interact with thecallback list500. Thecallback list500 may display one ormore data messages501. For example, the message(s) displayed may be the data message that was received while the user was unavailable and any other aggregate data that may be useful for the device user. In some embodiments, themobile device101 may be able to automatically generate smart reply messages that can be sent as a reply to the received data messages or voice messages (voicemails, missed calls) and displayed in the message(s)501 display. In addition to the displaying the assigned andsorted message501 thecallback list500 may also display a tentative, scheduledcallback time502 determined by thecallback integration engine201. Also present in this embodiment, are options to delete504 and move503 the scheduled callback from the callback list. For example, amobile device101 user may decide to send an email in response to the second item on thecallback list500 instead of proceeding with a scheduled callback, in this instance the user may choose to delete504 the callback from the callback list. Another option themobile device101 user may choose is to move503 the listed callback items. This may be done by selecting amove503 button on one callback list item and then selecting anothermove503 button on a different callback list item facilitating a swap of callback list items in the hierarchy. Another way to move503 callback list items may be by drag-and-dropping callback items. In this way, one callback item may be moved without having to swap places with another callback item. If an automated smart reply message has be generated and displayed in the callback list, themobile device101 user may review and amend the smart reply message within the callback list and then select an option to send the smart reply message. After a smart reply message has been sent, the user may choose to delete or maintain the callback item associated with the smart reply. Amobile device101 user may choose to confirm505 thecallback list500 without making any changes, or the user can make alterations (move, delete, swap, etc.) to thecallback list500 and then confirm505 it. If a confirmation is made without any alterations, thecallback integration engine201 may execute thecallback list500 at the displayed scheduledtimes502. If alterations are made to thecallback list500, then thecallback integration engine201 may need to reschedule callback items based upon the alterations implemented by themobile device101 user and then execute the scheduled callbacks.
• FIG.6 is a user interface diagram illustrating an exemplaryincoming call600 on amobile device101 with integrated callback features, according to an embodiment. According to the embodiment, amobile device101 with integrated callback features provided by acallback integration engine201 may natively incorporate callback functionality into standard functions such as receiving a phone (or voice over Internet protocol, “VoIP”) call, email (as described below with reference toFIG.7), or other functionality. When acall600 is received, the native answer prompt may present the user with the usual options to answer601 or decline603 the call, with normal functionality (generally, either answering the call and starting an interaction, or declining the call and sending it to voicemail or an automated prompt).
• Callback integration engine201 may present an additional prompt to setup acallback602, either through OS-based software integration with the callback integration engine operating at theOS layer220, or as a base-level firmware feature that is natively recognized and exposed by the OS while thecallback integration engine201 operates at thefirmware level320. This addedoption602 may be used to automatically request or schedule a callback, for example by providing a message to the caller requesting they call back at a predetermined time (for example, based on known availability from a user's on-device calendar), or by engaging with acloud callback platform110 to automatically arrange a callback that connects both participants. Whether or not thecallback prompt602 is presented may be configurable, such as by incorporating trust lists or zones that determine what callers may be eligible for a callback (similar to a “favorite contacts” list that may be able to call the user even when a do-not-disturb feature is enabled), or context-based configuration such as to provide a callback prompt when the user is in a meeting or otherwise scheduled as “busy” in their calendar, or when the user is driving.
• FIG.7 is a user interface diagram illustrating anexemplary email application700 operating on amobile device101 with integrated callback features, according to an embodiment. According to the embodiment, an email message or contact may be augmented with callback functionality provided bycallback integration engine201, presented as anoption701 to setup an automated callback from within anemail application700. This feature may incorporate information from the current email such as atopic703,timing704, orcontact information702, as well as any additional device, context, user, or other information that may be relevant (for example, calendar information or other emails outside the current message or thread), and cause a callback object to be generated. This callback object may then be used to automatically schedule and execute a callback by contacting theparticipants702 of the email (and it should be appreciated that this need not be limited to two participants, and may be used to create automated conference calls), and then bridge the individual calls to each participant to complete the interaction by connecting the participants together into a single call.
• Callbacks may be scheduled according to a variety of criteria, including (but not limited to) user availability as determined from preconfigured settings or known context (for example, calendar or email information such as invitation responses or verbal commitments in messages that may not have been separately entered into a calendar), user activity based on device information such as network or sensor data (such as if the device is paired to a car audio system, indicating the user is driving, or if there is significant accelerometer data that might indicate the user is in the middle of an exercise activity). Callbacks may then be scheduled to occur when the user is available or no longer indisposed, and may also incorporate availability on the part of the caller by (for example) providing them with a selection of callback options to from which to select a specific callback time. In addition to providing a callback selection on an incoming call prompt, the callback function may be exposed in other areas throughout the device's OS and applications, such as from within voicemail messages (to setup an automated callback with the caller that left the voicemail), social media apps (to setup automated callbacks with other users), or potentially any application operating on the device101 (such as to setup an automated callback for technical support). This integrated callback operation may be consistent throughout the device's software, providing a native user experience that blends seamlessly with the other features and elements of the device's operating system and applications.
• To generate, schedule, and execute callbacks, a callback object is created on the user'sdevice101 to represent the callback information such as scheduling, context information, user and caller information, and any additional data pertinent to the callback (for example, related interactions such as previous calls or emails exchanged with the caller, or a known call intent based on available information from the user and/or caller, such as email transcripts or voicemail messages). This callback object may be created and maintained on-device, operating locally within thecallback integration engine201 or within an application or feature of theoperating system220 of the user'sdevice101, enabling full callback functionality regardless of any connection to, or availability of, acloud callback platform110. In other arrangements, callback objects may be cloud-based to provide a centralized or software-as-a-service (SaaS) operation mode, for example to provide tiered or subscription-based callback functionalities offered by a callback cloud through handling of callback objects on behalf of users.
• FIG.8 is a method diagram illustrating an exemplary incoming call flow, according to an embodiment. According to the embodiment, a call is received801 at amobile device101. Acallback integration engine201 then analyzes the information available for thecall802, such as (for example, including but not limited to) caller ID, caller and recipient time zones, or whether the caller is a member of a trust zone in the user's settings or contact information (or a trust zone not configured by the user, such as a corporate trust zone for coworkers and colleagues). If the caller is eligible (that is, if they have sufficient trust or if their call is determined to be valuable to the user, such as a call from a technical support number for a company the user recently contacted), a callback token is associated with the call data803. If the user is untrusted, such as a blocked user or a suspected “spam” number, the call may be passed to the OS layer unmodified804. When received by the OS layer, the call is then displayed as an incoming notification as usual805, with a callback token (if available) used to place an additional callback button within the interface using the native OS and call notification user experience (UX) design. This provides an integrated automated callback functionality that is transparently incorporated into device features in a manner consistent with the operating system's UX and familiar to the user, blending the new functionality with the rest of the device features. It should be further appreciated that this mode of operation may function whether thecallback integration engine201 is operating at thefirmware320 orOS220 level, as many mobile operating systems such as ANDROID™ and IOS™ allow for applications to integrate with communication features such as phone dialers and incoming call notification prompts.
• FIG.9 is a method diagram illustrating an exemplary callback workflow once a user selects a callback for an incoming call, according to an embodiment. Initially, a callback button is selected901 by the user, either from an incoming call that they wish to defer to an automated callback or from an email for which they wish to automatically setup a callback, or any other interaction, application, or location where an integrated callback button may be present. A callback object is instantiated902, using acallback integration engine201, which is an object with data fields representing the various parts of callback data for the user and any other callback participants (for example, the caller if a callback button was pressed on an incoming call notification, or other individuals participating in an email thread if the callback button was pressed from within an email message), and any related information such as what scheduled times may be possible for such a callback to take place. This callback object is then stored and maintained by thecallback integration engine903, updating information when necessary such as to accommodate changes in scheduling or a participant indicating that they will be unavailable during the selected callback time. This may result in modifying the existing callback to reschedule it, and when the conditions for the callback are met (scheduled time arrives, users are available, or any other conditions that may have been set), thecallback integration engine201 initiates a call to each participant904 and then bridges them into asingle call905 where they may interact. This provides automated connection of multiple individuals as needed, without requiring any participant to initiate the call or remember scheduling information, as the entire process is handled “behind the scenes” by thecallback integration engine201.
• FIG.10 is a method diagram illustrating an exemplary callback prioritization workflow once a data or voice message is received, according to an embodiment. Initially, a data message or voice message (voicemail) may be received1001 by acomputing device101 wherein the user of the device is unable to respond to the received messages. In some embodiments, the voice message may be transcribed into a data message1002 for simpler processing by NLP algorithms. The data messages are then assigned to alobby construct1003 based upon one or more characteristics associated with the data message. The lobby constructs are groups of data messages that have similar or corresponding characteristics. For each assigned data message in a lobby, thecallback prioritization engine202 may retrieve andaggregate data1004 related to the assigned data message. Aggregate data may include previous text-based communications which have been identified to be related to the assigned data message. Identifying related information may be conducted using the one or more characteristics of the assigned data message. The assigned data message and any aggregate data may be input into one or more machine learning algorithms1005, such as, for example, a natural language processing neural network, to analyze the data message and its aggregate data1006. The analyses of the data message may provide, determine, derive, or infer context data (e.g., intent, sentiment, subject matter, etc.) which can provide more information about the received data messages. The results of the analyses may be used by thecallback prioritization engine202 to compute a priority score for the assigned data messages1007. Using the computed priority score, a callback list may be generated1008 which provides a hierarchal schedule of callback items that may be displayed to a user device for review and revision if necessary.
• FIG.11 is a block diagram illustrating an exemplary system architecture for an exemplarymobile device101 embodiment with an enhancedcallback prioritization engine1100 operating at the operating system level and connected to acloud callback platform110, according to an embodiment. According to the embodiment,callback cloud110 may receive requests via a plurality of communications networks such as a public switched telephone network (PSTN)103 or theInternet102. These requests may comprise a variety of communication and interaction types, for example including (but not limited to) voice calls over a telephone line, video calls over a network connection, or live text-based chat such as web chat or short message service (SMS) texting viaPSTN103. Such communications networks may be connected to a plurality ofmobile devices101 such as a user's smartphone or similar mobile device, according to the particular architecture of communication network involved.Mobile devices101 may be connected to respective communications networks via a variety of means, which may include telephone dialers, VOIP telecommunications services, web browser applications, SMS text messaging services, or other telephony or data communications services. It will be appreciated by one having ordinary skill in the art that such means of communication are exemplary, and many alternative means are possible and becoming possible in the art, any of which may be utilized as an element ofsystem100 according to the invention.
• When needed (for example, when a user manually requests a callback or when configured settings determine that a callback is needed), a user'smobile device101 creates a session with acloud callback platform110 with a profile manager111, resulting in a callback being required. Profile manager111 receives initial requests to connect tocallback cloud110, and forwards relevant user profile information to acallback manager113, which may further request environmental context data from anenvironment analyzer112. Environmental context data may include (for example, and not limited to) recorded information about when a user may be suspected to be driving or commuting from work (as may be determined from device information, such as whether a “do not disturb while driving” feature has been enabled, or if themobile device101 is currently connected to a car audio system), or if the user may be busy for another reason, such as if they are running or working out (as may be determined from device sensor data), for example, and may be parsed from online profiles or online textual data, using anenvironment analyzer112.
• Acallback manager113 centrally manages all callback data, creating a callback object which may be used to manage the data for a particular callback, and communicates with aninteraction manager114 which handles requests to make calls and bridge calls, which go out to amedia server115 which actually makes the calls as requested. In this way, themedia server115 may be altered in the manner in which it makes and bridges calls when directed, but thecallback manager113 does not need to adjust itself, due to going through an intermediary component, theinteraction manager114, as an interface between the two. Amedia server115, when directed, may place calls and send messages, emails, or connect voice over IP (“VoIP”) calls and video calls, to users over aPSTN103 or theInternet102.Callback manager113 may work with a user's profile as managed by a profile manager111, with environmental context from anenvironment analyzer112 as well as (if provided) EWT information for any callback recipients (for example, contact center agents with the appropriate skills to address the callback requestor's needs, or online tech support agents to respond to chat requests), to determine an appropriate callback time for the two users (a callback requestor and a callback recipient), interfacing with aninteraction manager114 to physically place and bridge the calls with amedia server115.
• Thecloud callback platform110 may further include an application programming interface (API)manager116 and acallback prioritization engine117, according to an embodiment. TheAPI manager116 may be used to expose or connect with software related to a callback requester and a callback recipient. For example,API manager116 may be used to connect with email, messaging, telephone, or other voice communication applications present on a mobile device in order to retrieve text-based data related to emails, text messages such as SMS and from other messaging applications, or voice data such as voicemails. Voice data such as, for example, a voice mail may be transcribed into a text-based data message by thecloud callback platform110. The retrieved text-based data messages may be sent to aprioritization engine117 which may identify one or more characteristics of the text-based data messages. Theprioritization engine117 may use the environment analyzer to parse or analyze the retrieved text-based data messages. Analysis of the text-based data messages may be used to determine or infer a variety of metrics pertaining to a callback requester and at least one callback recipient, such as (but not limited to) relationships, intent, subject matter, and sentiment. Based on the analysis of the text-based data messages, thecallback prioritization engine117 may produce priority score for each text-based data message which can be used to order lobbies (callback groups) in order of priority.
• According to the embodiment, amobile device101 comprises several hardware and software components operating at various levels to provide various functions. At the mostbasic hardware level210, the physical hardware of themobile device101 may comprise a processor andmemory211 that provide standard computing device functionality as described in detail below, with reference toFIGS.14-17 (grouped together here for the sake of clarity, it should be appreciated that processor/memory may or may not be part of the same hardware component, such as a system-on-a-chip), abaseband processor212 that manages radio-based communication functions such as cellular or Wi-Fi connectivity, as well as any of a number ofhardware sensors213 such as (for example, including but not limited to) an accelerometer to detect or measure device movement, gyroscope for detecting or measuring device orientation, barometer for measuring ambient environment conditions, or global positioning system (GPS) receiver for geolocating the device.
• An operating system (OS)220 comprises the main software operating onmobile device101 and providing various software-based functions such as software applications and communications.Various software applications221 operating within (for example in an application layer not shown here but as is commonly used in various computing devices according to the standard OSI process model of computing systems) on themobile device101 may expose and provide access to, or interaction with, various hardware and sensor features such as to enable a user to view or calibrate sensor readings.
• According to the embodiment, acallback integration engine201 may comprise local (that is, operating on the mobile device) functionality similar to a cloud-basedcallback manager113 described above, and may operate as an application or feature at theoperating system level220, running at a similar privilege level and having similar access to hardware and software functions asother applications221 operating onmobile device101. This enables easy installation or removal ofcallback integration engine201, as it may be readily distributed via similar methods as any other software application (for example, via an application store or similar download portal). However, this may restrict the capabilities ofcallback integration engine201, as it will have only indirect access to hardware features (as it will only be able to access whatever features are exposed by the operating system) and in many cases may be “sandboxed” and prevented from directly accessing other software or information on the mobile device201 (for example, in a sandboxed software environment such as APPLE IOS™ or similar).
• According to the embodiment, an enhancedcallback prioritization engine1100 may comprise local (that is, operating on the mobile device) functionality similar to a cloud-basedprioritization engine117 described above, and may operate as an application or feature at theoperating system level220, running at a similar privilege level and having similar access to hardware and software functions asother applications221 operating onmobile device101. The enhancedcallback prioritization engine1100 may be included within the same application as thecallback integration engine201 or it may be packaged as its own application with built in functionality to communicate with and share data with a separatecallback integration engine201. This enables easy installation or removal of an enhancedcallback prioritization engine1100, as it may be readily distributed via similar methods as any other software application (for example, via an application store or similar download portal).
• Thecallback integration engine201 may intercept (e.g., receive, hook, access, or otherwise obtain) voice messages (i.e., voicemail) and text-based data messages such as (but not limited to) short message service (SMS), emails, and group chats sent to themobile device101 and route them to an enhancedcallback prioritization engine1100 as a callback object associated with the intercepted data or voice messages. In an embodiment, themobile device101 may transcribe voice messages into text-based data message for further processing. Each text-based data message and voice message (or its transcription) may have one or more characteristics, such as indications of source, destination, data message subject matter, end user identifiers or account identifiers.Application221 data associated with the data message is aggregated and linked to the callback object. For example, a first email sent to amobile device101 may be intercepted and then the enhancedcallback prioritization engine1100 can identify and retrieve other emails, text messages, group chats, financial transactions, social media interactions fromapplication221 on thephone101 and associated with the first email in order to provide more context for the first email message. In this example, other communications exchanged between the first email recipient and the first email sender (such as an email thread, text messages, or voicemails) may be aggregated and added to the callback object associated with the first email. The aggregated data may then be fed into one or more machine learning algorithms that analyze the aggregated data and can generate for each data message a priority score. The priority score may be used to create a hierarchy or priority list of callback objects such that each callback may be scheduled and executed in order of highest priority to lowest priority.
• According to another embodiment, an exemplarymobile device101 with acallback integration engine201 and an enhancedcallback prioritization engine1100 can operate at the device firmware level, according to an embodiment. According to the embodiment, a firmware level encompasses low-level program code that operates “below” theoperating system220, operating directly on hardware components of mobile101, and comprises such features as a baseband controller that controls features ofbaseband processor212 and the operation of which is fully transparent to the user (that is, a user ofmobile device101 does not see or interact with firmware, and many users may not even be aware of its existence or capabilities).
• According to the embodiment, acallback integration engine201 and an enhancedcallback prioritization engine1100 may operate as a firmware component that directly interfaces with hardware features ofmobile device101, enabling direct control as needed while exposing desired functionality to theoperating system220 in a transparent manner (that is, the operating system may only know that a feature is available, and may not be able to determine that the feature is provided by thecallback integration engine201 orcallback prioritization engine202 rather than, for example, the baseband controller). This enables more direct control over device functions, such as call and data message routing and hardware information such as sensor readings (as may be used in callback workflow processing), and allows thecallback integration engine201 to intercept incoming calls and radio information (e.g., data messages) before it reaches the operating system, so data may be manipulated and additional features may be integrated prior to presentation to the OS. This also prevents accidental removal ofcallback integration engine201 and an enhancedcallback prioritization engine1100, as a user may have no control over device firmware and anything operating at or below theOS level220 would inherently have insufficient access to modify firmware. This further enablescallback integration engine201 and an enhancedcallback prioritization engine1100 to access hardware features that may be restricted or completely prevented when operating within theOS layer220, for example hardware sensor or geolocation information that may be incorporated into a callback workflow.
• According to another embodiment, wherein the enhancedcallback prioritization engine1100 runs at theoperating system220 level, an agent1101 may be included with the enhancedcallback prioritization engine1100 that is provided with root access and privileges in order to monitor the mobile device and access all files and folders needed to achieve to retrieval of third-party app data for use in the machine learning analysis.
• FIG.12 is a block diagram illustrating an exemplary system architecture for an enhancedcallback prioritization engine1100, according to an embodiment. According to the embodiment, an enhancedcallback prioritization engine1100 may receive a callback object with a data message comprising a characteristic from acallback integration engine201. The data message can have part of sessions that include asynchronous text, voice message, or data message (e.g., not a live voice conversation such as a telephone call) communications between and end user and a computing device (e.g., mobile device, contact center agent [e.g., a person, bot, or combination thereof], etc.). In some examples, an enhancedcallback prioritization engine1100 can receive voice messages that have been sent as audio messages (voicemail) or converted into text messages. The sessions can occur continuously in a conversational manner, or can be intermittent, e.g., lasting over hours, days, or weeks with dormant periods between communications. The enhancedcallback prioritization engine1100 receiving data messages with characteristics can then find other relevant application data using the information in the callback object.
• The enhancedcallback prioritization engine1100 may identify and aggregate other data related to the data message. For example, if an assigned data message is part of an ongoing session (such as an email thread, text, or group messaging conversation) then adata aggregator402 may retrieve the other messages contained within the session and append or link the other messages to the assigned data message. Additionally, if the assigned data message is a standalone message (i.e., sent with no connection to a prior session) then the enhancedcallback prioritization engine1100 may scan for and identify other communications that may be related to the assigned data message. Other communications may be identified as related based upon one or more characteristics of the assigned data message. For example, an end user may receive an email asking for a response of some sort from a coworker which may be related to previous communications exchanged between the end user and the coworker, but not contained within the same thread as the previous communications. In this example, the enhancedcallback prioritization engine1100 may be able to identify the previous communications as related to the original email asking for a response and append or link them to the assigned data messages generated from the response email. As another example, an end usermobile device101 may receive a voicemail from an individual or entity (business) asking the end user to callback when available. The voicemail may be transcribed to a text data message and have one or more characteristics, such as the phone number or other identifier associated with the voicemail, and thedata aggregator1201 can use the voicemail characteristics to identify an email (or other data message) that may be associated with the voicemail. For example, an employee of a company may send to a second employee a voicemail requesting a callback and further, the two employees may have been emailing or text messaging each other earlier in the day or week. These emails or text messages may be linked to the transcribed voicemail by thedata aggregator1201. In this way thedata aggregator1201 may identify and aggregate application data correlating with a data message.
• Social media apps may also be associated with a callback object. For example, two competing callback requests, both scoring the same in a callback queue—may be resolved by determining recent social media interactions. If social media interaction is higher between one of the callback requesters and the user, then that callback request can be placed in a higher priority than the other. Social media history may comprise likes, mentions, thumbs-up/down, or other means the service uses to indicate sentiment. If the caller is a moderator or administrator on a forum or other outlet where the user is a member or has recent activity, then higher priority can be given over regular or non-members.
• Financial apps may also be used to determine if a callback requester is associated with an invoice, pay request, or other financial transaction. Financial app integration may inform themachine learning algorithms1202 to prioritize callbacks based on transactions amounts, e.g., a first callback requester who sent the user a large monetary sum is given priority over a second callback requester that has sent a small amount of funds.
• The assigned data message and any linked and aggregated data may then be sent as input to a machine learning (ML)engine1202 for processing and analysis. TheML engine1202 may comprise one or more machine learning algorithms such as, for example, a natural language processing (NLP) neural network. The NLP may be used for text classification functions including, but not limited to, sentiment analysis, topic categorization, language recognition, and intent analysis. The NLP may be used to perform sentiment analysis on the assigned data message and any available aggregated data linked to the data message. Sentiment analysis may be used to determine if the content of a data message (and its aggregate data) is positive, negative, or neutral. This could be used to classify data messages to better respond to inquiries or requests. The NLP may be used to determine urgency. Identifying keywords and sentiments that provide insight and context into the callback request. The NLP may be used for topic categorization to determine which category a data message belongs to, essentially this constitutes a robust, automatic method of tagging and organizing the content of a data message or its aggregate data. The NLP may be used to understand the intent of a data message or its aggregate data by determining the underlying intention of a given data message. In addition, the NLP may be used to find and identify relationships between a caller and callee, or a message recipient and a message sender, or between and among a group of people (i.e., multiple people in an email thread or group text message). Using the identified relationships, an enhancedcallback prioritization engine1100 orcallback integration engine201 may generate personalized smart replies based on the identified relationships and other information (intent, sentiment, urgency, topic) derived from the NLP. The relationships may be used to determine the priority of an assigned data message. Additionally, NLP can also be used to determine urgency of a callback. Given certain keywords such as “it's important”, “right away”, and “urgent” can be used to give some callback requests higher priority than others.
• In an embodiment, theML engine1202 may also comprise one or more audio machine learning algorithms which may be used for audio classification and speech recognition functions. Audio classification ML algorithms may be used when a voice message (voicemail) is received bymobile device101 or by an enhancedcallback prioritization engine1100. The audio algorithm may be used to identify a speaker by taking a snippet of the voice message to create a unique identifier, like a voice fingerprint. Then the fingerprint may be input into a neural network that can match the unique identifier to a known or previously heard voice. This may allow an enhancedcallback prioritization engine1100 to identify callers who may be calling from a new device or using a new identifier that may not be known by thecallback integration engine201. Furthermore, an audio ML algorithm may be utilized for speech recognition functions such as voice to text technology (i.e., transcribing a voicemail into a text data message) or to provide voice assistant interaction (e.g., in a contact center environment whereby a customer may communicate with a bot or script voice assistant).
• There are various benefits of having aML engine1202 operating locally on a mobile device compared to operating in a cloud environment which the mobile device may connect to. Operating theML engine1202 on amobile device101 protects against low latency. As most ML algorithms may be operating in real-time, network speed may be too slow to transmit a live data feed from the device to thecloud callback platform110, and then back to the device. Furthermore, running theML engine1202 locally on amobile device101 removes the risk of low or no network connections halting the ML processes. In amobile device101 theML engine1202 may still work when run on device, even if low or no connectivity. Operating theML engine1202 locally on amobile device101 facilitates privacy over the mobile device user's data. ML needs data to build a useful model and to make useful inferences derived from the model, however, some data may be sensitive (e.g., personal identifying information, regulated data). The enhancedcallback prioritization engine1100 can use and secure data on the device, without having to transmit any sensitive data.
• TheML engine1202 may forward the results of the analyses (e.g., inferences, intent, sentiment, urgency, topic, etc.) of the various ML algorithms to aprioritization module1203. Theprioritization module1202 may then use the results of the analyses (e.g., inferences, intent, sentiment, urgency, topic, etc.) to generate a priority score for each data message. For example, if multiple assigned messages within a construct have been identified to belonging to the same topic, then the most recent data message pertaining to the shared topic may be given a higher priority score and thus the callback associated with that data message would be executed before the previously received data messages. For example, a cable company service agent may have twenty five data messages, and of those data messages twenty are about a loss of cable service. In this example, theML engine1202 may be able to identify that each of those twenty data messages originated from the same area code and therefore there must be some service disruption affecting that area code. Theprioritization module1203 can use this information to move the first of the twenty people up in priority so that the problem with their cable service may be addressed. If the cable service agent is able to resolve the issue for the first of twenty data messages relating to cable service loss, then the enhancedcallback prioritization engine1100 may be able to send an automated message to the other of the customers associated with the remaining nineteen data messages providing information about the loss of cable service. For example, an automated message may be sent which says, “Is your call about loss of cable service? If so, we are aware, and your service should be restored soon. Would you like to request a callback?”. If the receiver of the automated message replies with a callback request, then the data message may remain in the queue, otherwise the data message will be removed from the callback queue.
• Intent and sentiment analysis results may be used to determine priority as well. For example, the sentiment of one data message may be determined to be positive, whereas the sentiment of another data message may be urgent, or negative. A data message with an urgent or negative sentiment may be given a higher priority score than a data message with a positive sentiment in order to account for the urgency of the data message, or to quickly respond to the negative sentiment data message in order to change sentiment of the data message sender to neutral or positive. Additionally, the intent of a data message may be used to determine the priority of response. For example, a data message may indicate that there was no intent for a callback or response, whereas another data message may have an inferred intent for a callback. In this case, the data message with the inferred intent for a callback will be given a higher priority score than the data message with little to no inferred callback intent.
• Based upon the priority score of each data message, theprioritization module404 may reorder the callback queue in a hierarchal manner where the highest priority data message is at the top of the queue and the other data messages are sorted in descending order of priority. Theprioritization module1203 may also produce a callback list based upon the sorted queue. The callback list may include the original data message, any aggregate data, and data contained within the callback object associated with each data message. For example, the callback list may display the original data message, a scheduled callback time, and the name of the callback recipient. The callback list may be sent to the mobile device101 (or any computing device) for display and interaction by the mobile device user.
• According to various embodiments,machine learning engine1202 may further comprise one or more machine learning algorithms configured to identify callback recipients that are most-likely to result in a desired outcome (e.g., business outcome, for example, duration of interactions), or an outcome that is most-positive. The one or more machine learning algorithms (referred to herein as an “outcome model”) may use at least one of callback object data, data message information, determined context and urgency output from prioritization model(s), callback recipient data, and aggregated application data as inputs in order to analyze potential recipients for a callback. In an embodiment, the callback recipients are contact center agents and the one or more machine learning models may analyze each contact center agent based on a plurality of callback recipient data which can include, but is not limited to, agent skills, QA scores, scheduling, previous interactions, outcome record, and/or the like. In some embodiments, the outcome model analyzes callback recipients according to one or more selected business outcomes, such as duration of interactions, customer satisfaction rating, agent idle time, etc. For example, if a contact center has an abundance of idle agents, an administrator can instruct model generating module to generate a ranking model that routes interactions to agents in a manner that minimizes agent idle time.
• According to some embodiments,prioritization module1203 computes priority scores for a plurality of callback objects based on machine learning analysis and then creates a priority list of the plurality of callback objects based on the computed priority scores. The priority list may comprise the original data message, any aggregate data, and output context and urgency information from one or more machine learning algorithms.Prioritization module1203 may produce an outcome score that ranks callback recipients according to their predicted interaction outcomes for each data message contained in the priority list. Each callback object associated with each data message of the priority list may be assigned the “best” callback recipient based on the outcome score and/or ranking. In some embodiments, a priority list need not be generated prior to outcome model performing its analysis, furthermore, in such embodiments outcome model may begin analyzing potential callback recipients as soon as a callback object is received from thecallback integration engine201 or other component capable of creating and transmitting callback objects.
• Enhancedcallback prioritization engine1100 may test the efficacy of an outcome model (and/or prioritization models) in order to ensure that the outcome model properly assigns callback recipients. In some implementations, the model is tested to ensure properly assigns callback recipients according to desired business outcomes. In some implementations, enhancedcallback prioritization engine1100 performs one or more simulations of a recipient ranking process and analyzes the results.Prioritization module1203 may test an outcome model by monitoring routed interactions and determining how well business outcomes are achieved, using actual recipient profiles, test profiles, or a combination thereof.Prioritization module1203 may then analyze the resulting recipient-requestor pairings to determine whether incoming data messages are being assigned to the recipients that are most likely to lead to a positive outcome. The actual outcome data may be used to refine scoring and the predictive capabilities of the outcome model over time with continuous iterations of training, testing, and monitoring outcomes.
• In some implementations, outcome model can be further configured (e.g., trained and tested) for predictive outcome scoring on a per-requestor basis. Requestor-specific information may be used as input data into the outcome model. Requestor specific information may be used as inputs into outcome model in conjunction with callback recipient data and the determined context and urgency associated with a data message or set of data messages. For example, a specific user may be easy/hard to please and this can change the scoring for various recipients. If a particular user is especially hard to please, then an contact center agent may be selected as a callback recipient based on their ability to manage difficult customers or that have a lot of experience in a particular context category associated with data message associated with the difficult to please user. Requestor specific information can include requestor preferences. For example, a user may have preferred agents or agent qualities. Such preferences can include a preferred agent (i.e., callback recipient) language, location, scheduling, and prior interaction outcomes. Requestor specific information can be referred to as requestor parameters/attributes/factors interchangeably herein. In these embodiments, each callback requestor's outcome scoring is different because the requestor itself is a contextual factor that is incorporated into the scoring.
• Prioritization module1203 can generate a callback list that may include the original data message, any aggregate data, and data contained within the callback object associated with each data message, as well as an assigned callback recipient, when applicable.
• An additional anticipated feature of the instant invention is incentivized on-device callback management. This is where rewards are offered to the mobile device user for handling a callback in a specified manner. This could be a specific time—e.g., “call back within this window for a bonus!”, or “if you call back, use code ABC for a discount; only valid for twenty-four hours” Specific channels could be incentivized such as asking a user to reach out on a specific platform. Rewards could be in the form of in-game rewards, money, discounts, add-ons to their mobile service plan, priority handling on their next initiated interaction, or discounted or free service provided by the caller—e.g., “get in touch with us via our streaming channel for a free month!”.
• Incentivized on-device callback management may be implemented by using NLP to determine the incentives from the data message and adding a function to each callback object meeting the criteria of the data message's incentive. For example, if a user is given an incentive to call back within 24 hours to receive 20% off of becoming a new member of a service, a message or notification may be generated on the users device alerting them of the offer and providing a link to the specific callback object on the callback list.
• FIG.13 is a method diagram illustrating several app integrations for enhancing callback prioritization. This diagram illustrates features of thedata aggregator1201 using APIs and other functions (e.g., screen monitoring, administrative access, etc.) to retrieve information from the mobile device in order to provide context to callback request scoring and ordering. When a characteristic is known about thecallback request1301—e.g., caller name, location, business, demographics, subject matter, etc., thedata aggregator1201 retrieves, from the applications in the mobile device, anycommunications1302 from the same caller, business, subject matter, etc. If enough data points from themachine learning1202 suggest a communication or interaction from an external application is relevant, that communication or interaction may be used with the original callback request analysis to provide some context with regards to sentiment, urgency, relationships, and intent. The higher the machine learning confidence is for the sentiment, urgency, relationships, and intent of the external application data in relation to the callback request, the higher the value of a prioritization score will be1302. This same analysis and scoring continue forsocial media interactions1303/1303a,financial interactions1304/1304a, andgaming interactions1305/1305a. Sensors and sensor history may be used, apart from or alongside of calendar or to-do apps to determine if the caller and the user have been in the same area or met physically recently1306/1306a. For example, after a recent trip to a specific business in China, an American business executive may receive a follow-up call from the business clerk. The callback will reach a high priority based on the travel data and nature of the user's business.
• Example sentiments, relationships, and intents comprise, but are not limited to scheduled events involving the caller and recent events involving caller determined by GPS history or shared events or calendars. Whether a caller is a team member, party member, or a competitor. Whether the caller a developer, moderator, player, or otherwise involved in a game the user plays. Recent interactions on social media may be analyzed such as likes, mentions, thumbs-up/down, or some means the service uses to indicate sentiment. The caller may a moderator or administrator on a forum or other outlet where the user is a member or has recent activity. It could be determined if a caller owes the user money, the user owes the caller money, or the caller and user have exchanged money recently.
• Eachprioritization score1302a,1303a,1304a,1305a,1306ais then summed1307 and used to place the callback object in the appropriate place in thecallback list1308.Prioritization scores1302a,1303a,1304a,1305a,1306amay fall on any scale or range desired and may be normalized as seen fit, evident to those with ordinary skill in the art.
• FIG.14 is message flow diagram illustrating an exemplary exchange of data messages between components of a cloud callback platform with active callback prioritization, according to an embodiment. According to the embodiment, the embodiment comprises at least acallback integration engine201, adata aggregator1201, anML engine1202, and aprioritization module1203. Additional components which may be present but are not shown in the illustration can include one or more databases configured to store callback recipient (e.g., contact center agent) profiles and/or information and requestor (e.g., customer or user) profiles and/or information.Callback integration engine201 can receive one or more data messages such as email, text message, instant message, etc. Responsive to receiving a data messagecallback integration engine201 may create a callback object which can comprise information such as the data message and user (e.g., individual who received the data message) information. The callback object can be sent1405 todata aggregator1201 which can request a plurality of application data from a plurality of software applications operating on the user device. Data aggregator receives, retrieves, or otherwise obtains the plurality of app data and aggregates the data and associates it with the callback object.Data aggregator1201 can then send1410 the callback object and aggregated data toML engine1202.ML engine1202 may comprise one or more machine learning models configured to analyze input data to produce as output predictions, scores, and/or classifications. In some embodiments, ML engine may comprise a prioritization model which analyzes input data such as a data message, a callback object, and any available aggregated data, and produces as output a determined context and urgency associated with the received data message.ML engine1202 send its output context andurgency analysis data1415 toprioritization module1203 which can use the output analysis data to compute a priority scored for each data message of a plurality of data messages.
• Prioritization module1203 can send thepriority list data1420 back toML engine1202. The priority list may comprise at least a list of data messages and for each data message its determined context and/or urgency data. AtML engine1202, a second set of one or more machine learning models configured to analyze input data to produce as output predictions, scores, and/or classifications. In some embodiments, ML engine may comprise an outcome model which analyzes input data such as callback recipient information and context category information, and produces as output a outcome score associated with a callback recipient for a given context category. ML engine may receive fromdata aggregator1201 the callback object andcallback recipient data1425. AtML engine1202 the recipient data and context data contained with the priority list may be fed as input in to the outcome model to produce interaction outcome analysis which may be an outcome score, or used to compute an outcome score for a callback recipient. The analysis of the outcome model can be sent toprioritization module1203 was interaction outcome analysis data1430.Prioritization module1203 may then use the interaction outcome analysis to assign outcome scores which can be applied to generate a prioritizedcallback list1435 wherein callback recipients are assigned to each callback requestor based on a predicted outcome. In some embodiments, the outcome may be a business outcome such as minimizing agent idle time or maximizing customer satisfaction.
• FIG.15 is message flow diagram illustrating an exemplary exchange of data messages between components of a cloud callback platform with active callback prioritization and requestor-specific outcome scoring, according to an embodiment. According to the embodiment, the embodiment comprises at least acallback integration engine201, adata aggregator1201, anML engine1202, and aprioritization module1203. Additional components which may be present but are not shown in the illustration can include one or more databases configured to store callback recipient (e.g., contact center agent) profiles and/or information and requestor (e.g., customer or user) profiles and/or information.Callback integration engine201 can receive one or more data messages such as email, text message, instant message, etc. Responsive to receiving a data messagecallback integration engine201 may create a callback object which can comprise information such as the data message and user (e.g., individual who received the data message) information. The callback object can be sent1505 todata aggregator1201 which can request a plurality of application data from a plurality of software applications operating on the user device. Data aggregator receives, retrieves, or otherwise obtains the plurality of app data and aggregates the data and associates it with the callback object.Data aggregator1201 can then send1510 the callback object and aggregated data toML engine1202.ML engine1202 may comprise one or more machine learning models configured to analyze input data to produce as output predictions, scores, and/or classifications. In some embodiments, ML engine may comprise a prioritization model which analyzes input data such as a data message, a callback object, and any available aggregated data, and produces as output a determined context and urgency associated with the received data message.ML engine1202 send its output context andurgency analysis data1515 toprioritization module1203 which can use the output analysis data to compute a priority scored for each data message of a plurality of data messages.
• Prioritization module1203 can send thepriority list data1520 back toML engine1202. The priority list may comprise at least a list of data messages and for each data message its determined context and/or urgency data. AtML engine1202, a second set of one or more machine learning models configured to analyze input data to produce as output predictions, scores, and/or classifications. In some embodiments, ML engine may comprise an outcome model which analyzes input data such as callback recipient information and context category information, and produces as output a outcome score associated with a callback recipient for a given context category. In this embodiment, the outcome model is further configured to determine outcome scores on a per-requestor basis by using requestor-specific information as a context input factor in model analyses. ML engine may receive fromdata aggregator1201 the callback object andcallback recipient data1525. Additionally, in this embodiment,data aggregator1201 may receive, retrieve, or otherwise obtain a plurality of requestor-specific information, such as by retrieving a requestor profile stored in a database.Data aggregator1201 can send thecallback requestor data1530 to ML engine for analysis. AtML engine1202 the recipient data, context data contained with the priority list, and requestor-specific data may be fed as input in to the outcome model to produce interaction outcome analysis which may be an outcome score, or used to compute an outcome score for a callback recipient. The analysis of the outcome model can be sent toprioritization module1203 as requestor-specificoutcome analysis data1535.Prioritization module1203 may then use the interaction outcome analysis to assign outcome scores which can be applied to generate a prioritizedcallback list1540 wherein callback recipients are assigned to a specific callback requestor based on the requestor-specific outcome. In some embodiments, the outcome may be a business outcome such as minimizing agent idle time or maximizing customer satisfaction.
• FIG.16 is a flow diagram illustrating an exemplary method for training and testing a machine learning model, according to an embodiment. According to the embodiment, the machine learning model is trained to determine an outcome score associated with a callback recipient of a plurality of callback recipients and a context category. In some implementations, the callback recipients are contact center agents. According to the embodiment, the process begins at1601 whenmachine learning engine1202 receives a plurality of input data and creates a training dataset and test dataset using pre-processed data selected from the plurality of input data. The plurality of input data can include, but is not limited to, interaction histories, agents, and context categories. Agent data may comprise a plurality of assigned and/or defined attributes. Interaction histories may be specific to an agent, a context category, or a combination thereof. At1602 the training dataset is used to train one or more machine learning algorithms in order to construct a machine learning model which can determine an outcome score for an agent for each of a plurality of context categories. The output of the machine learning model may be observed and analyzed during model training to determine model behavior at1603. It is during this operation that the test dataset may be used to test the efficacy of the trained model. In some embodiments, the test dataset comprises simulated data. In some embodiments, the input data comprises synthetic data. If the model is not behaving as expected then the model parameters can be tuned at1604. Examples of model parameters that can be tuned/optimized between iterations of model training can include the following (non-limiting): weights, support vectors, coefficients, learning rate, number of nodes, optimization algorithm, activation function, loss function, number of epochs, number of clusters, pooling size, batch size, etc. The tuned parameters can be applied to the next round of model training at1602.
• At1603, if the model behavior is sufficient, as determined by some metric such as error rate, false positive rate, and/or the like, then the trained machine learning model may be deployed at1605. Once deployed, a model may be used in a production environment wherein the model receives as input real data (i.e., data not drawn from the training input set) and determines an outcome score based at least on the real data. The outcome score can be used to assign an agent to a callback requestor based in part on the context of a data message corresponding to a priority list. Once each entry on a priority list has been assigned an agent (i.e., callback recipient) a callback list may be generated and used to organize a callback queue. At1606 the interaction between an agent and a callback requestor is monitored and the interaction outcome is recorded or otherwise determined. The interaction outcome represents the “real” or actual outcome and can be applied as feedback at1602 in order to refine the scoring capabilities of the machine learning model. In some embodiments, the machine learning model is an outcome model.
Hardware Architecture
• Generally, the techniques disclosed herein may be implemented on hardware or a combination of software and hardware. For example, they may be implemented in an operating system kernel, in a separate user process, in a library package bound into network applications, on a specially constructed machine, on an application-specific integrated circuit (“ASIC”), or on a network interface card.
• Software/hardware hybrid implementations of at least some of the aspects disclosed herein may be implemented on a programmable network-resident machine (which should be understood to include intermittently connected network-aware machines) selectively activated or reconfigured by a computer program stored in memory. Such network devices may have multiple network interfaces that may be configured or designed to utilize different types of network communication protocols. A general architecture for some of these machines may be described herein in order to illustrate one or more exemplary means by which a given unit of functionality may be implemented. According to specific aspects, at least some of the features or functionalities of the various aspects disclosed herein may be implemented on one or more general-purpose computers associated with one or more networks, such as for example an end-user computer system, a client computer, a network server or other server system, a mobile computing device (e.g., tablet computing device, mobile phone, smartphone, laptop, or other appropriate computing device), a consumer electronic device, a music player, or any other suitable electronic device, router, switch, or other suitable device, or any combination thereof. In at least some aspects, at least some of the features or functionalities of the various aspects disclosed herein may be implemented in one or more virtualized computing environments (e.g., network computing clouds, virtual machines hosted on one or more physical computing machines, or other appropriate virtual environments).
• Referring now toFIG.17, there is shown a block diagram depicting anexemplary computing device10 suitable for implementing at least a portion of the features or functionalities disclosed herein.Computing device10 may be, for example, any one of the computing machines listed in the previous paragraph, or indeed any other electronic device capable of executing software- or hardware-based instructions according to one or more programs stored in memory.Computing device10 may be configured to communicate with a plurality of other computing devices, such as clients or servers, over communications networks such as a wide area network a metropolitan area network, a local area network, a wireless network, the Internet, or any other network, using known protocols for such communication, whether wireless or wired.
• In one embodiment,computing device10 includes one or more central processing units (CPU)12, one ormore interfaces15, and one or more busses14 (such as a peripheral component interconnect (PCI) bus). When acting under the control of appropriate software or firmware,CPU12 may be responsible for implementing specific functions associated with the functions of a specifically configured computing device or machine. For example, in at least one embodiment, acomputing device10 may be configured or designed to function as a serversystem utilizing CPU12,local memory11 and/orremote memory16, and interface(s)15. In at least one embodiment,CPU12 may be caused to perform one or more of the different types of functions and/or operations under the control of software modules or components, which for example, may include an operating system and any appropriate applications software, drivers, and the like.
• CPU12 may include one ormore processors13 such as, for example, a processor from one of the Intel, ARM, Qualcomm, and AMD families of microprocessors. In some embodiments,processors13 may include specially designed hardware such as application-specific integrated circuits (ASICs), electrically erasable programmable read-only memories (EEPROMs), field-programmable gate arrays (FPGAs), and so forth, for controlling operations ofcomputing device10. In a specific embodiment, a local memory11 (such as non-volatile random access memory (RAM) and/or read-only memory (ROM), including for example one or more levels of cached memory) may also form part ofCPU12. However, there are many different ways in which memory may be coupled tosystem10.Memory11 may be used for a variety of purposes such as, for example, caching and/or storing data, programming instructions, and the like. It should be further appreciated thatCPU12 may be one of a variety of system-on-a-chip (SOC) type hardware that may include additional hardware such as memory or graphics processing chips, such as a QUALCOMM SNAPDRAGON™ or SAMSUNG EXYNOS™ CPU as are becoming increasingly common in the art, such as for use in mobile devices or integrated devices.
• As used herein, the term “processor” is not limited merely to those integrated circuits referred to in the art as a processor, a mobile processor, or a microprocessor, but broadly refers to a microcontroller, a microcomputer, a programmable logic controller, an application-specific integrated circuit, and any other programmable circuit.
• In one embodiment, interfaces15 are provided as network interface cards (NICs). Generally, NICs control the sending and receiving of data packets over a computer network; other types ofinterfaces15 may for example support other peripherals used withcomputing device10. Among the interfaces that may be provided are Ethernet interfaces, frame relay interfaces, cable interfaces, DSL interfaces, token ring interfaces, graphics interfaces, and the like. In addition, various types of interfaces may be provided such as, for example, universal serial bus (USB), Serial, Ethernet, FIREWIRE™, THUNDERBOLT™, PCI, parallel, radio frequency (RF), BLUETOOTH™, near-field communications (e.g., using near-field magnetics), 802.11 (Wi-Fi), frame relay, TCP/IP, ISDN, fast Ethernet interfaces, Gigabit Ethernet interfaces, Serial ATA (SATA) or external SATA (ESATA) interfaces, high-definition multimedia interface (HDMI), digital visual interface (DVI), analog or digital audio interfaces, asynchronous transfer mode (ATM) interfaces, high-speed serial interface (HSSI) interfaces, Point of Sale (POS) interfaces, fiber data distributed interfaces (FDDIs), and the like. Generally,such interfaces15 may include physical ports appropriate for communication with appropriate media. In some cases, they may also include an independent processor (such as a dedicated audio or video processor, as is common in the art for high-fidelity A/V hardware interfaces) and, in some instances, volatile and/or non-volatile memory (e.g., RAM).
• Although the system shown inFIG.17 illustrates one specific architecture for acomputing device10 for implementing one or more of the inventions described herein, it is by no means the only device architecture on which at least a portion of the features and techniques described herein may be implemented. For example, architectures having one or any number ofprocessors13 may be used, andsuch processors13 may be present in a single device or distributed among any number of devices. In one embodiment, asingle processor13 handles communications as well as routing computations, while in other embodiments a separate dedicated communications processor may be provided. In various embodiments, different types of features or functionalities may be implemented in a system according to the invention that includes a client device (such as a tablet device or smartphone running client software) and server systems (such as a server system described in more detail below).
• Regardless of network device configuration, the system of the present invention may employ one or more memories or memory modules (such as, for example,remote memory block16 and local memory11) configured to store data, program instructions for the general-purpose network operations, or other information relating to the functionality of the embodiments described herein (or any combinations of the above). Program instructions may control execution of or comprise an operating system and/or one or more applications, for example.Memory16 ormemories11,16 may also be configured to store data structures, configuration data, encryption data, historical system operations information, or any other specific or generic non-program information described herein.
• Because such information and program instructions may be employed to implement one or more systems or methods described herein, at least some network device embodiments may include nontransitory machine-readable storage media, which, for example, may be configured or designed to store program instructions, state information, and the like for performing various operations described herein. Examples of such nontransitory machine-readable storage media include, but are not limited to, magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROM disks; magneto-optical media such as optical disks, and hardware devices that are specially configured to store and perform program instructions, such as read-only memory devices (ROM), flash memory (as is common in mobile devices and integrated systems), solid state drives (SSD) and “hybrid SSD” storage drives that may combine physical components of solid state and hard disk drives in a single hardware device (as are becoming increasingly common in the art with regard to personal computers), memristor memory, random access memory (RAM), and the like. It should be appreciated that such storage means may be integral and non-removable (such as RAM hardware modules that may be soldered onto a motherboard or otherwise integrated into an electronic device), or they may be removable such as swappable flash memory modules (such as “thumb drives” or other removable media designed for rapidly exchanging physical storage devices), “hot-swappable” hard disk drives or solid state drives, removable optical storage discs, or other such removable media, and that such integral and removable storage media may be utilized interchangeably. Examples of program instructions include both object code, such as may be produced by a compiler, machine code, such as may be produced by an assembler or a linker, byte code, such as may be generated by for example a JAVA™ compiler and may be executed using a Java virtual machine or equivalent, or files containing higher level code that may be executed by the computer using an interpreter (for example, scripts written in Python, Perl, Ruby, Groovy, or any other scripting language).
• In some embodiments, systems according to the present invention may be implemented on a standalone computing system. Referring now toFIG.18, there is shown a block diagram depicting a typical exemplary architecture of one or more embodiments or components thereof on a standalone computing system.Computing device20 includesprocessors21 that may run software that carry out one or more functions or applications of embodiments of the invention, such as for example aclient application24.Processors21 may carry out computing instructions under control of anoperating system22 such as, for example, a version of MICROSOFT WINDOWS™ operating system, APPLE OSX™ or iOS™ operating systems, some variety of the Linux operating system, ANDROID™ operating system, or the like. In many cases, one or more sharedservices23 may be operable insystem20, and may be useful for providing common services toclient applications24.Services23 may for example be WINDOWS™ services, user-space common services in a Linux environment, or any other type of common service architecture used withoperating system21.Input devices28 may be of any type suitable for receiving user input, including for example a keyboard, touchscreen, microphone (for example, for voice input), mouse, touchpad, trackball, or any combination thereof.Output devices27 may be of any type suitable for providing output to one or more users, whether remote or local tosystem20, and may include for example one or more screens for visual output, speakers, printers, or any combination thereof.Memory25 may be random-access memory having any structure and architecture known in the art, for use byprocessors21, for example to run software.Storage devices26 may be any magnetic, optical, mechanical, memristor, or electrical storage device for storage of data in digital form (such as those described above, referring toFIG.17). Examples ofstorage devices26 include flash memory, magnetic hard drive, CD-ROM, and/or the like.
• In some embodiments, systems of the present invention may be implemented on a distributed computing network, such as one having any number of clients and/or servers. Referring now toFIG.19, there is shown a block diagram depicting anexemplary architecture30 for implementing at least a portion of a system according to an embodiment of the invention on a distributed computing network. According to the embodiment, any number ofclients33 may be provided. Eachclient33 may run software for implementing client-side portions of the present invention; clients may comprise asystem20 such as that illustrated inFIG.18. In addition, any number ofservers32 may be provided for handling requests received from one ormore clients33.Clients33 andservers32 may communicate with one another via one or moreelectronic networks31, which may be in various embodiments any of the Internet, a wide area network, a mobile telephony network (such as CDMA or GSM cellular networks), a wireless network (such as Wi-Fi, WiMAX, LTE, and so forth), or a local area network (or indeed any network topology known in the art; the invention does not prefer any one network topology over any other).Networks31 may be implemented using any known network protocols, including for example wired and/or wireless protocols.
• In addition, in some embodiments,servers32 may callexternal services37 when needed to obtain additional information, or to refer to additional data concerning a particular call. Communications withexternal services37 may take place, for example, via one ormore networks31. In various embodiments,external services37 may comprise web-enabled services or functionality related to or installed on the hardware device itself. For example, in an embodiment whereclient applications24 are implemented on a smartphone or other electronic device,client applications24 may obtain information stored in aserver system32 in the cloud or on anexternal service37 deployed on one or more of a particular enterprise's or user's premises.
• In some embodiments of the invention,clients33 or servers32 (or both) may make use of one or more specialized services or appliances that may be deployed locally or remotely across one ormore networks31. For example, one ormore databases34 may be used or referred to by one or more embodiments of the invention. It should be understood by one having ordinary skill in the art thatdatabases34 may be arranged in a wide variety of architectures and using a wide variety of data access and manipulation means. For example, in various embodiments one ormore databases34 may comprise a relational database system using a structured query language (SQL), while others may comprise an alternative data storage technology such as those referred to in the art as “NoSQL” (for example, HADOOP CASSANDRA™, GOOGLE BIGTABLE™, and so forth). In some embodiments, variant database architectures such as column-oriented databases, in-memory databases, clustered databases, distributed databases, or even flat file data repositories may be used according to the invention. It will be appreciated by one having ordinary skill in the art that any combination of known or future database technologies may be used as appropriate, unless a specific database technology or a specific arrangement of components is specified for a particular embodiment herein. Moreover, it should be appreciated that the term “database” as used herein may refer to a physical database machine, a cluster of machines acting as a single database system, or a logical database within an overall database management system. Unless a specific meaning is specified for a given use of the term “database”, it should be construed to mean any of these senses of the word, all of which are understood as a plain meaning of the term “database” by those having ordinary skill in the art.
• Similarly, most embodiments of the invention may make use of one ormore security systems36 andconfiguration systems35. Security and configuration management are common information technology (IT) and web functions, and some amount of each are generally associated with any IT or web systems. It should be understood by one having ordinary skill in the art that any configuration or security subsystems known in the art now or in the future may be used in conjunction with embodiments of the invention without limitation, unless aspecific security36 orconfiguration system35 or approach is specifically required by the description of any specific embodiment.
• FIG.20 shows an exemplary overview of acomputer system40 as may be used in any of the various locations throughout the system. It is exemplary of any computer that may execute code to process data. Various modifications and changes may be made tocomputer system40 without departing from the broader scope of the system and method disclosed herein. Central processor unit (CPU)41 is connected tobus42, to which bus is also connectedmemory43,nonvolatile memory44,display47, input/output (I/O)unit48, and network interface card (NIC)53. I/O unit48 may, typically, be connected to peripherals such as akeyboard49, pointingdevice50,hard disk52, real-time clock51, acamera57, and other peripheral devices.NIC53 connects to network54, which may be the Internet or a local network, which local network may or may not have connections to the Internet. The system may be connected to other computing devices through the network via arouter55, wirelesslocal area network56, or any other network connection. Also shown as part ofsystem40 ispower supply unit45 connected, in this example, to a main alternating current (AC)supply46. Not shown are batteries that could be present, and many other devices and modifications that are well known but are not applicable to the specific novel functions of the current system and method disclosed herein. It should be appreciated that some or all components illustrated may be combined, such as in various integrated applications, for example Qualcomm or Samsung system-on-a-chip (SOC) devices, or whenever it may be appropriate to combine multiple capabilities or functions into a single hardware device (for instance, in mobile devices such as smartphones, video game consoles, in-vehicle computer systems such as navigation or multimedia systems in automobiles, or other integrated hardware devices).
• In various embodiments, functionality for implementing systems or methods of the present invention may be distributed among any number of client and/or server components. For example, various software modules may be implemented for performing various functions in connection with the present invention, and such modules may be variously implemented to run on server and/or client components.
• The skilled person will be aware of a range of possible modifications of the various embodiments described above. Accordingly, the present invention is defined by the claims and their equivalents.

Claims (20)

What is claimed is:

1. A mobile device with active callback prioritization, comprising:

a processor, a memory, and a plurality of programming instructions stored in the memory and operable on the processor;

a callback integration engine comprising a subset of the plurality of programming instructions that, when operating on the processor, cause the processor to:

produce a callback object in memory comprising information associated with a received data or voice message; and

send the callback object to an enhanced callback prioritization engine; and

the enhanced callback prioritization engine comprising a subset of the plurality of programming instructions that, when operating on the processor, cause the processor to:

receive the callback object from the callback integration engine;

retrieve and aggregate application data related to the data or voice message;

retrieve callback recipient data for a callback recipient;

use the callback object and the aggregated application data as inputs into a prioritization model, wherein the prioritization model analyzes the callback object's information and the aggregated application data to determine the context and urgency associated with the data or voice message;

for each callback object, compute a priority score based at least upon the results of the analysis;

use the determined context and urgency and the retrieved callback recipient data as inputs into one or more machine learning algorithms, wherein the machine learning algorithms analyze the context and urgency and the retrieved callback recipient data to determine an interaction outcome associated with the data or voice message; and

use the computed priority score, the callback object data, and the data or voice message to generate a callback list.

2. The mobile device ofclaim 1, wherein the enhanced callback prioritization engine is further configured to:

retrieve and aggregate requestor-specific data;

use the aggregated requestor-specific data as an additional input into the one or more machine learning algorithms, wherein the machine learning algorithms analyze the context and urgency data, the retrieved callback recipient data, and the aggregated requestor-specific data to determine a requestor-specific outcome associated with the data or voice message;

for each callback recipient, compute a requestor-specific outcome score based at least upon the results of the analysis; and

assign a callback recipient to and the callback object based at least upon the requestor-specific outcome score.

3. The mobile device ofclaim 1, wherein the callback recipient data comprises agents skills, quality assurance scores, scheduling information, historical interactions, spoken languages, certifications, interaction outcomes, and location information.

4. The mobile device ofclaim 2, wherein the requestor-specific data comprises requestor preferences.

5. The mobile device ofclaim 4, wherein the requestor preferences comprise preferred agents or agent qualities.

6. The mobile device ofclaim 1, wherein the application data is retrieved using application programming interfaces.

7. The mobile device ofclaim 1, wherein the callback list comprises a smart reply message.

8. The mobile device ofclaim 1, wherein the callback integration engine receives a user confirmed callback list and executes the confirmed callback items on the list.

9. The mobile device ofclaim 1, wherein the application data comprises data from communication, social media, financial, gaming, and productivity applications.

10. The mobile device ofclaim 1, wherein the machine learning algorithms comprise natural language processing.

11. A method for active callback prioritization, comprising the steps of:

producing a callback object in memory comprising information associated with a received data or voice message received;

sending the callback object to an enhanced callback prioritization engine;

receiving the callback object from the callback integration engine;

retrieving and aggregate application data related to the data or voice message;

retrieving callback recipient data for a callback recipient;

using the callback object and the aggregated application data as inputs into a prioritization model, wherein the prioritization model analyzes the callback object's information and the aggregated application data to determine the context and urgency associated with the data or voice message;

for each callback object, computing a priority score based at least upon the results of the analysis;

using the determined context and urgency and the retrieved callback recipient data as inputs into one or more machine learning algorithms, wherein the machine learning algorithms analyze the context and urgency and the retrieved callback recipient data to determine an interaction outcome associated with the data or voice message; and

using the computed priority score, the callback object data, and the data or voice message to generate a callback list.

12. The method device ofclaim 11, further comprising the steps of:

retrieving and aggregate requestor-specific data;

using the aggregated requestor-specific data as an additional input into the one or more machine learning algorithms, wherein the machine learning algorithms analyze the context and urgency data, the retrieved callback recipient data, and the aggregated requestor-specific data to determine a requestor-specific outcome associated with the data or voice message;

for each callback recipient, computing a requestor-specific outcome score based at least upon the results of the analysis; and

assigning a callback recipient to and the callback object based at least upon the requestor-specific outcome score.

13. The method ofclaim 11, wherein the callback recipient data comprises agents skills, quality assurance scores, scheduling information, historical interactions, spoken languages, certifications, interaction outcomes, and location information.

14. The method ofclaim 12, wherein the requestor-specific data comprises requestor preferences.

15. The method ofclaim 14, wherein the requestor preferences comprise preferred agents or agent qualities.

16. The method ofclaim 11, wherein the application data is retrieved using application programming interfaces.

17. The method ofclaim 11, wherein the callback list comprises a smart reply message.

18. The method ofclaim 11, wherein the callback integration engine receives a user confirmed callback list and executes the confirmed callback items on the list.

19. The method ofclaim 11, wherein the application data comprises data from communication, social media, financial, gaming, and productivity applications.

20. The method ofclaim 11, wherein the machine learning algorithms comprise natural language processing.

Images