Hi, I'm Omkar.

I build models, and then I find out whether they actually helped.

I spent six years as a data scientist, designing experiments and getting models into production. The one I'm still proud of is a triage system for radiologists that cut SLA breaches on urgent orders by 10%.

After that, four years cofounding startups building data science and AI tools. For the last two I've been building production AI agents, along with the eval harness that tells me whether they're any good.

What I like most is the part between the model and the decision: figuring out what's worth building, and how we'd know if it worked.

I'm currently looking for data science or ML roles.

If your team is wrestling with models, experiments, or whether an LLM system is really working — get in touch.