Technical roots with Product mindset

Most AI products don't fail because the model wasn't good enough. They fail because the product layer was not designed with enough rigor — the intent capture, the trust mechanics, the human-AI handoff. I've watched this from both sides.

I spent nearly a decade as an engineer building AI systems before they were called agentic: integrity systems at Meta tackling fraud and harmful content for 3.5 billion users, and financial infrastructure at Stripe moving money across borders at internet scale. Earlier in my career, I built cloud automation systems at Microsoft and pricing infrastructure at GoDaddy.

That engineering foundation was built on a deliberate educational bet: a Master's in Computer Science from Carnegie Mellon, specializing in AI and Machine Learning, on top of degrees in Computer Science and Economics from BITS Pilani. The economics degree is what taught me that systems have incentives, and that products fail when the incentive structure isn't designed as carefully as the architecture.

When I crossed over to product, I didn't leave that engineering and analytical lens behind. Today I lead product for Issue Planner at CodeRabbit — a 0→1 bet on the hardest problem in AI-assisted development: taking a human's ambiguous intent and turning it into a structured coding plan an AI agent can execute.