// backend systems that hold up in production
I build backends that turn raw data into something useful: aggregating job feeds, streaming video, and replaying market history. Clean APIs, sensible architecture, no surprises when traffic shows up.
I'm a software engineer who gravitates toward the backend where data gets ingested, transformed, and served without breaking. That's meant a job-aggregation platform with vector-based matching, a video service that transcodes and segments uploads for smooth streaming, and a C++ engine for backtesting trading strategies.
I care about systems that are correct first and fast second, and I like the whole path from a raw data source to something a person can actually use.
Building a Kafka-based weather alert system that streams events at 100/sec through a JSON-driven rule engine, fronted by a React dashboard with severity filtering. Added role-based access control and audit logging built to the NIST 800-53 standard.
A job-hunting platform that pulls fresh listings daily from multiple job APIs, matches them to each user, and serves a personalized dashboard, with a built-in AI cover-letter writer. Users, jobs, and their vector embeddings all live in PostgreSQL for semantic matching.
A streaming backend where users upload raw video that's transcoded with FFmpeg, chunked into ordered segments, and served for smooth playback. Segments live in AWS S3 with metadata tracked in PostgreSQL.
A backtesting engine that replays recorded market data so you can run trading strategies (buy-and-hold and beyond) and see how they'd have performed. The C++ core handles the simulation; a companion Python tool visualizes and compares results.
I'm open to software engineering roles and always happy to talk shop. The fastest way to reach me is email.