I build Python services and data tools, with a focus on reliable behaviour and clear tests. My work includes webhook delivery, dataset checks and language-model evaluation. I've also contributed accepted fixes to Black and urllib3.

Portfolio   /   LinkedIn   /   Merged contributions

For backend work, start with RelayForge. For algorithms and data handling, try SplitCheck. Pressroom shows a complete web application.

RelayForge A webhook service that keeps track of deliveries, signs requests and retries when a receiver is unavailable. Workers can recover unfinished jobs after a crash. Python FastAPI PostgreSQL Run the demo · Read the database tests

SplitCheck A command-line tool that finds exact text matches and shared groups across ML dataset splits. It indexes training data, then checks evaluation rows one at a time using hash-based lookups. Python Data structures pytest Algorithm and memory benchmark · Streaming tests

Authors draft, editors review, and only approved revisions go live. Includes revision history, scheduled publishing, a REST API and PostgreSQL concurrency tests.

Run it locally · Architecture · Concurrency tests

My MSc research project examines language-model responses across multi-turn conversations. It includes an evaluation workflow, annotation tools and offline verification, with demonstration data kept separate from private research evidence.

Python · SQL · PostgreSQL · FastAPI · Django · pytest · Docker · GitHub Actions

I also work with JavaScript and TypeScript for web interfaces and their tests.

Want to discuss a project or contribution? Message me on LinkedIn.