I work at the intersection of Data, AI, and Product. Builder is probably the more honest job title.
Lately, I’ve been dangerously skipping permissions (a lot!) to push the envelope on what AI Agents can do reliably.
I currently work on the Threads App, building the ranking systems that power the Notifications stack.
Before that, I was at Meta Superintelligence Labs (MSL). At MSL, I worked on understanding the Open Source LLM ecosystem, and how to make LLMs work for real use cases; how to make them reliable, performant, and actually deliver user value. In practice, this meant digging into pre-training data, combing through evals, and reading more model outputs than any person reasonably should.
And like most people in tech at some point, I worked on ads; part of the team that built Click-to-WhatsApp and Lead Ads for Business Messaging from a nascent idea into products at scale.
Before Meta, I spent a few years at FLYR Labs, an airline pricing startup, bringing machine learning into legacy revenue systems. Despite three years of pricing data, I still pay as much for a flight as anyone else.
I’m a quantitative economist by training. That means I learned to run linear regressions the hard way and am always the first person to ask if something is correlation or causation.
I’m most energized when I'm working on 0→1 projects; happiest when I’m learning, building, and shaping ambiguous problems into real products that work.
You can reach me at vignesh.ananth@gmail.com.