Invoice AI Auditor
An LLM pipeline that ingests invoice PDFs, extracts every line item, reconciles them against the ledger, and auto-flags financial discrepancies — replacing a manual review process and saving the ops team hours per cycle.
Six years of schema-to-UI ownership. I scaled DockerCon's virtual platform to 70,000+ concurrent users, and I'm now building production LLM pipelines with FastAPI & Claude.
An LLM pipeline that ingests invoice PDFs, extracts every line item, reconciles them against the ledger, and auto-flags financial discrepancies — replacing a manual review process and saving the ops team hours per cycle.
I don't hand off at the API boundary. Schema to UI — I take ownership of the whole vertical slice.
A working product in users' hands beats a perfect one in review. Bias toward shipping and learning.
Give me a problem and a deadline. I'll figure out the path, flag blockers early, and deliver.
I've kept 70k+ users online at once. I know when to optimise and when to keep it simple.
“Schema to UI, problem to ship — I do the whole vertical.”
I'm a full-stack engineer based in Hyderabad with six years of building and shipping products end to end. I've kept 70k+ users online simultaneously for Docker's flagship conference, built an LLM-powered audit tool at Kaamwork, and shipped side projects from idea to production solo.
These days I work across React, NestJS, and Python/FastAPI — increasingly at the intersection of AI and product. I care about clean schemas, fast feedback loops, and not handing off at the API boundary.
Open to remote full-time roles where I can take ownership, move fast, and build things that matter.
Open to remote full-time roles & interesting contracts. I reply within 24 hours.