Backend systems that don't fall over at 2 a.m.

I'm a backend engineer who's spent the last year across four different production codebases — enterprise data pipelines, decentralized health records, serverless AWS infrastructure, and AI-agent platforms, mostly figuring out how to keep them running.

Lately most of my work sits where backend systems meet LLMs: multi-agent architectures, retrieval pipelines, and the unglamorous plumbing auth, fallbacks, rate limits, that makes an AI product reliable enough to actually ship.

Outside of internships I build my own things ~ a CFO copilot, a backend scaffolding CLI, a multilingual healthcare voice assistant and compete in hackathons, a habit that's landed a few finalist placements along the way.