MLOPS: ML ENGINEERING BEST PRACTICES FROM THE TRENCHES
Date: Thursday, 31 October 2019
Time: 2:00PM
Location: TBD
DevOps tools for getting code reliably to production have proven to be effective in the software engineering world. Today, ML Engineers are working at the intersection of data science and software engineering, and can leverage DevOps best practices to streamline their workflow and delivery process. This is what MLOps is all about.
At Manifold, we've developed processes to help ML engineers work as an an integrated part of your development and production teams, helping you to be deliberate, disciplined, and coordinated in your deployment process. In this workshop, Sourav and Alex will walk you through some key learnings from using this Lean AI process. They’ll cover topics such as:
- Scaffolding a project with a consistent structure for ML
- Using Docker for a consistent runtime across the team and environments
- Cleanly separating ML configuration from ML source code
- Setting up an ML experiment tracking system
- Systems for rapid ML experimentation in the cloud
- Deploying ML seamlessly at production scale using Docker
