Hi, I'm Shivek.

I'm a full-stack software engineer working toward a focused goal: becoming an AI Software Engineer who can take a machine learning idea from notebook to production. I care about the unglamorous parts most people skip — evaluation, retrieval quality, error handling — because that's where AI products actually succeed or fail.

My background is in full-stack web development, which shapes how I approach AI work: models are only useful inside systems that are fast, observable, and built for real users. I spend my time building projects that combine both — RAG pipelines, model serving infrastructure, and automation agents — and writing about what I learn along the way.

Outside of shipping projects, I'm deepening my machine learning fundamentals and studying how production AI systems are designed at scale. This site is both a portfolio and a public log of that process.