Repository files navigation

“It doesn’t guess. It learns — and then it loops.”
A deterministic, self-refining insight engine that transforms raw operational data into structured, actionable intelligence.
No cloud dependencies. No black-box AI. Just Python, Flask, and clarity.

TheRPLE ingests structured metadata (CSV), identifies emerging patterns, filters noise, and outputs continuously improving predictions and insights.
Each cycle compounds analytical clarity — meaningthe more you use it, the smarter it gets.

Built for transparency, not mystery: every decision is logged, scored, and explainable.

• 📊CSV Ingestion & Normalization — feed any structured operational dataset
• 🔁Recursive Insight Loop — learns from every cycle, compounding precision
• 🧩Constructive Feedback Engine — outputs clear insights, not just numbers
• 🧠Memory Reservoir — tracks persisting vs. novel patterns over time
• 🪶Lightweight Deployment — pure Python + Flask, no external APIs or services

graph TDA[CSV Upload] --> B[Data Normalization]B --> C[Pattern Detection & Scoring]C --> D[Insight Generation]D --> E[Human Feedback + Memory Reservoir]E --> C

1. Upload a CSV file (any dataset withvalue_1,value_2, and optionalrisk columns).

2. The engine analyzes:

   • Trends
   • Correlation shifts
   • Anomalies
   • Risk alignments

3. It generatesinsight cards with:

   • Confidence
   • Novelty
   • Severity
   • Suggested actions
   • Status (🆕new / ♻️persisting)

4. The engine stores each insight in its memory reservoir for future comparison.

git clone https://github.com/<your-handle>/recursive-logic-engine.gitcd recursive-logic-enginepython3 -m venv .venvsource .venv/bin/activatepip install -r requirements.txtpython app.py

Then open your browser athttp://127.0.0.1:5000 and upload your dataset.

sample_input.csv

timestamp,value_1,value_2,risk2025-10-01,10,12,0.202025-10-02,11,12,0.212025-10-03,13,13,0.222025-10-04,15,14,0.252025-10-05,18,14,0.302025-10-06,16,13,0.282025-10-07,19,15,0.332025-10-08,21,16,0.362025-10-09,24,17,0.40

Each loop stores a hashed summary of every insight ininsight_memory.json.Future runs detect whether insights are:

• 🆕 New: unseen patterns
• ♻️ Persisting: confirmed patterns continuing across cycles

This creates a real-time feedback model that grows smarter with use.

• Python 3.9+
• Flask 3.x
• Pandas + NumPy
• JSON + Local Storage (no external API)
• SHA-256 integrity for insight memory

Joseph Wells📍 Indianapolis, IN📧joepwells95@gmail.com🔗Foxhunter Labs

• 🦊Foxhunter Pro — Human-Gated Reconnaissance & Ethical Autonomy System
• 🧬Enigma² — Safety & Kill-Switch Engine
• 🛰️Swarm — Deterministic Multi-Agent Coordination Framework

MIT License © 2025 Joseph WellsUse freely for educational and research purposes. Attribution required.

“Predictive clarity doesn’t just happen — it compounds.”