I build machine learning systems and the data infrastructure that keeps them fed.
I graduated from the College of Engineering, Pune in 2020 with a B.Tech in
Computer Science. An internship at Veritas shaped what I wanted to work on —
an impact-analysis engine that mapped the dependencies between microservices and
scored what would break if any one of them went down.
I spent the five years after graduating at Bitwise, working across several
clients. The first couple of years were data engineering — pipelines, warehouses
and the unglamorous plumbing that makes analytics trustworthy — and the last of
them were spent as a machine learning engineer for a payment technology client,
taking models from notebook to production. Along the way I finished a PG Diploma
in Data Science at IIIT Bangalore, which is where a long-standing interest in
deep learning turned into something I could actually ship.
I am now a data scientist at Medline India, building product
intelligence and search over the product catalogue with GraphRAG — pairing a
knowledge graph of how products relate to each other with LLM-based retrieval,
so that a search can follow those relationships rather than just match text.
Day to day that means Python and GraphRAG; Azure CosmosDB, ML Studio and
Microsoft Fabric; PyTorch and TensorFlow when there is a model to train, and a
good deal of time on knowledge graphs and retrieval quality. I care about
systems that are boring to operate and easy to reason about six months later.
This site is also where I write. Expect posts on applied AI, data science and
cloud engineering — the things I had to figure out the hard way, written down so
the next person does not have to.
