Senior full-stack engineer at thinkbridge, ~6 years in. I build and ship products end to end — Angular and TypeScript on the front, Node on the back — and I lead the team's AI workstream.
At thinkbridge I run the AI workstream: rolling Claude out across the engineering org and designing how agentic features get built into our products. I lead and architect this work rather than write the model code myself — my hands-on lane is full-stack product, not ML research.
On the side I ship small, public, fully automated tools:
govai-contracts — a free, searchable index of U.S. federal AI/ML contract awards. Astro + TypeScript, fed from USAspending.gov and SAM.gov, statically generates per-award/agency/vendor/NAICS pages, and refreshes every 6 hours via a GitHub Actions cron that auto-commits the data. It also ships Ask GovAI, a measured RAG over the dataset — BM25 retrieval → grounded, cited LLM synthesis with a refuse-to-invent guardrail and a reproducible eval (recall@8 = 92%; 69% BM25 baseline before the structured query router). Live: govai-contracts.nandanhegde1096.workers.dev
mcp-directory — a searchable directory of public Model Context Protocol servers, auto-updated on a schedule.
TypeScript, Angular, Node.js. Astro, GitHub Actions, Cloudflare Workers, AWS. Comfortable across the front end, the API layer, and the build-and-deploy pipeline.
Portfolio: nandanhegde1.github.io/portfolio · LinkedIn: linkedin.com/in/nandanhegde1096
Open to senior and lead roles in full-stack and AI product engineering, including relocation.