Moizz K

I'm a self-taught AI engineer based in Islamabad, Pakistan — currently completing a BSc in Data Science, but I didn't wait for a degree to start building. I've been designing and shipping production AI systems since before the coursework caught up. That bias toward execution hasn't changed.

My flagship project, DocuMind, is a multi-document RAG research assistant built entirely on my own — no team, no tutorial, no shortcuts. It handles real documents at scale, enforces mandatory source citations on every response, and runs on production infrastructure with LangGraph multi-agent routing and Pinecone Serverless retrieval. It's the kind of system I'd want to use myself: reliable, cited, and honest about what it doesn't know.

Today I work globally as a full-stack AI engineer, building RAG pipelines, AI agents, and production LLM systems for clients who need things that actually work — not just in demos, but under real load, with real data, maintained by real teams. My stack centers on LangChain, LangGraph, FastAPI, and React, with retrieval backends chosen per problem rather than per habit. I work async-first, communicate clearly, ship on time, and don't disappear after handoff.