DevOps/MLOps on feature platform

Background

The organization’s branch responsible for identity management was undergoing a technology modernization effort. As part of the effort they were looking to incorporate more machine learning capabilities into their solution. I was hired onto a team of developers and machine learning experts to help build a domain specific feature platform.

Responsibilities

• collaborate with the team to architect a solution, providing recommendations in many areas: infrastructure & compute scaling, configuration driven architecture, asset & feature versioning, deployment and release workflows
• own infrastructure, infrastructure automation, deployments, & change management
• write documentation and create architecture and user-journey diagrams
• collaborate on the core feature API & compute engine

Business Value

• quickly deploy infrastructure to unblock prototyping & testing efforts
• float between roles - DevOps, MLOps, developer - to ensure we hit important milestones
• help establish relationships with external teams working on similar problems to enable knowledge sharing & collaboration
• complimented for my ability to present very technical concepts & designs to non-technical stakeholders