Anil Muppalla
Engineering Manager, AI Agent and LLM Evaluation Infrastructure
San Francisco, California, United States · 1 connectionNot on Unlinked · Imported profile. Is this you? Claim it
Is this about you? Request removal
About
The trickiest part of building AI agents isn't the model, it's the harness and the evals. I lead two engineering teams at Spotify working on exactly that problem. I've worked the full personalization stack: real time data infrastructure, ML feature platforms, and now the agents and evaluation systems that sit on top of it. I founded the squad behind the feature store and behavioral data that Spotify's home screen runs on. These days I'm building the agents that automate how we test and fix personalization surfaces, along with the evaluation platform and experimentation systems they run on. I design them to remove the toil and keep engineers on the judgment calls. Defining and measuring "good" at scale is the hard part, and investing in that rigor early is a force multiplier. The same is true of growing people. Since 2022 I've taken engineers from early career into senior and staff scope. I operate as a player-coach and stay close to the frontline: I still write and review code every week, which is how you mentor from real experience. Building agents and measuring them is where most of my attention goes right now. The measurement half is the unsolved bit: how you score a system that
Experience
Apr 2012 to Mar 2015
Apr 2018 to Mar 2026
Education
Master of Science (MS), Computer Science
Master of Science (MS) Computer Science
Bachelor of Engineering - BE, Computer Science
Bachelor of Engineering - BE Computer Science