Demonstration repo for the 2021 Nashville Analytics Summit session "Get Your Model to Production Faster by Wrapping it in a Flask App"

There are currently three subdirectories in this repo:

• flask-ml-api is a bare bones Flask app that you could use as a starting point for wrapping your ML model in an API.
• diabetes-predictor is a working example of what a simple ML model might look like using flask-ml-api as a starting point.
• live-demo is the directory we'll start from for some live coding fun during the session; our goal will be to reproduce (more or less) what's in diabetes-predictor, discussing some of our choices and thinking along the way.

• Flask: Python framework for Web development
• joblib Python functions as pipeline jobs
• schedule: Human-readable scheduling

• Cookiecutter Data Science
• Google Site Reliability Engineering
• Quantifying Technical Debt