A Tutorial on Robust Machine Learning Deployment
A Tutorial on Robust Machine Learning Deployment


A hands-on tutorial for productionizing machine-learning models using robust open-source tools. This tutorial shows you how to go from a python scikit model, get REST API endpoint, test it for common deployment issues, containerize, and deploy it. This is performed using a new open-source package, DRUM, that moves beyond flask and takes advantage of NGINX and uWSGI for serving model in a production-grade manner.

This package provides support for a variety of modeling frameworks including: Keras, scikit learn, R, H2O, DataRobot, and more. The package also incorporates unit testing for common deployment issues. All of this is easy to containerize and even add monitoring agents.

Session Outline
Introduction to model deployment
A hands on session that will:
Take a python scikit model and get a REST API endpoint
Use this REST API point to build a simple app
Add DataRobot monitoring agents to track the health of the deployment

Background knowledge:
No background necessary, but will be using python code


Rajiv Shah is a data scientist at DataRobot, where his primary focus is helping customers improve their ability to make and implement predictions. Previously, Rajiv has been part of data science teams at Caterpillar and State Farm. He has worked on a variety of projects from a wide ranging set of areas including supply chain, sensor data, acturial ratings, and security projects. He has a PhD from the University of Illinois at Urbana-Champaign.