Personal Edition · No Paywall · Mostly True
Open to opportunities — Agentic AI & LLM Systems — see credentials →
I strive to use both the logical and creative sides of the brain. Started as a UI/UX designer, went on to architect mobile apps in React Native and iOS, moved into web, and now I'm on the agentic AI path — building LLM agents, RAG systems, and the evaluation tools that tell you if they actually work. Still building for the human using the product, not just the system behind it.
Off the clock I write on Substack and Medium.
Engineering Lead — Sentient Labs
Built LLM-powered agent systems, agent success measurement, and agent evaluation pipelines for production reasoning agents.
Software Development Lead — Zeta (PayZapp / HDFC Bank)
Led architecture for PayZapp's React Native app and a 5-engineer team delivering PixelCard for HDFC Bank.
Technical Lead — mfine
Led mobile architecture and production deployment for an AI-powered healthcare platform.
Agent Reliability Lab
A LangGraph support agent with a live reviewer console and an offline harness that scores it at scale.
RAG Retrieval Strategy
An educational RAG debugger — inspect chunking, embedding, dense search, BM25, and hybrid fusion at every stage.
Why Retrieval Becomes the Hardest Part of RAG
Production RAG systems succeed or fail on retrieval quality, not LLM choice.
AI Infrastructure Investment Cycle (2023 → 2030)
Understanding the bottlenecks, capital rotation, and next AI opportunities.