Welcome to the Data Science Cookbook!

A resource to help data scientists go from problem to prototype

» Use the navigation bar at the top of this page
» Go to about for more details on what the cookbook is
» Go to ingredients
» Go to recipes
» Go to resources for we like
» Go to code examples for specific problems (computer vision, NLP, etc)

Author’s note: This is still a work in progress…

What is this?

An open-source “cookbook” to enable bottom-up data science in DoD units at the edge of the formation. Help small teams get from 0 → 1. Check out the about page for more details on the cookbook.

How do I use this cookbook?

Like any cookbook, flip to the recipe you need. We tried to make each subsection stand on its own and be consumable on its own. If you don’t understand a particular ingredient in the recipe, use the ingredients section to learn more.

Why did you make this?

Enablers of ML in DoD—infrastructure, funding, acquisition—are centralized. Problems, problem solving, and prototyping—are decentralized. We want to help merge top-down enablers with bottom-up problem solving

Centralized and Decentralized
Source: Made by Authors

How does it work?


Data scientists, designers, and product managers with experience in DoD (AI Task Force, DIU, Kessel Run, Fort Meade, SOCOM, Harbinger and more)


Operations Research Analysts, junior data analysts, data scientists, recent graduates of universities and grad school programs, bootcamp grads, product managers.

Where do I go from here?

» Take me to the about section
» Show me the ingredients
» Take me to the recipes
» Show me example code of machine learning in action


These views and ideas are the authors’ own and do not represent the views of the DoD, the Services, or any DoD organizations