I'm Peter. Nice to meet ya.

I largely help social sector organizations get their data into a shape where machine learning can be valuable. Much of this work ends up ondrivendata.org, where you canjoin a competition to help these organizations, learn from interesting data, try new methods, and make friends that care about impact. Here are some cool recent ones:

• Detect building heights and angles in satellite imagery
• Predict who is likely to get a flu vaccination
• Identify memes that are "hateful" based on both the text and the image
• Forecast geomagnetic solar storms

Competitions are great, but not every problem is a good fit, so our team of data scientists and software engineers also works with organizations directly to analyze data, build data systems, setup pipelines, train machine learning models, and design and deploy solutions. Check outDrivenData Labs to learn more. There Iwrite case studies,publish on our blog, andmaintain our open source work.

You can find me working on open source projects that are tools for data scientists and engineers using Python. I particularly care about reproducible data science and machine learning and AI ethics.

See below for the projects I regularly contribute to!

PinnedLoading

1. drivendataorg/cookiecutter-data-sciencedrivendataorg/cookiecutter-data-sciencePublic

   A logical, reasonably standardized, but flexible project structure for doing and sharing data science work.

   Python 8.6k 2.5k

2. drivendataorg/deondrivendataorg/deonPublic

   A command line tool to easily add an ethics checklist to your data science projects.

   Python 292 52

3. drivendataorg/cloudpathlibdrivendataorg/cloudpathlibPublic

   Python pathlib-style classes for cloud storage services such as Amazon S3, Azure Blob Storage, and Google Cloud Storage.

   Python 524 66

4. drivendataorg/nbautoexportdrivendataorg/nbautoexportPublic

   Automatically export Jupyter notebooks to various file formats (.py, .html, and more) on save.

   Python 75 9

5. drivendataorg/zambadrivendataorg/zambaPublic

   A Python package for identifying 42 kinds of animals, training custom models, and estimating distance from camera trap videos and images

   Python 125 29

6. drivendataorg/pandas-pathdrivendataorg/pandas-pathPublic

   Use pathlib syntax to easily work with Pandas series containing file paths.

   Python 69 4