Parbhat Kapila

AI Full-Stack Engineer who ships production RAG systems and LLM-powered products. I specialize in the intersection of modern web development and AI—building semantic search engines, document intelligence platforms, and streaming LLM interfaces that handle real user traffic.

I've built end-to-end AI applications from scratch: RAG architectures with pgvector, multi-provider LLM orchestration with fallback handling, real-time document processing pipelines, and SaaS billing with Stripe. My approach is serverless-first (Next.js Edge, Vercel) with Redis caching, concurrent processing, and cost-aware API design.

I don't just integrate OpenAI and call it done. I think about token limits, streaming UX, rate limiting, error recovery, and building systems that stay within budget at scale. I'm comfortable working across the full stack—PostgreSQL query optimization, React streaming UIs, webhook idempotency, and everything in between.

Looking to join an early-stage AI startup (seed to Series A) where I can own meaningful parts of the product and work directly with founders to ship features users want to pay for.

Stack: Next.js 14+, TypeScript, React, Node.js, PostgreSQL/pgvector, Redis, LangChain, tRPC, OpenAI/Anthropic/Gemini APIs, Vercel/AWS, Docker, Stripe.