From friction to shipped product.

PoCo — a reference library, polished and completed.

PoCo (Polish & Complete) turns an inconsistent Zotero or EndNote export into a clean, consistent reference library — without ever touching your original file. It standardises journal names, DOIs, author names, page ranges and title casing, completes missing publication metadata, and flags duplicate DOIs and retraction notices, then hands you a polished copy plus a full audit trail.

The engine is deterministic by design: no LLM, no hallucination. Every proposed change is a scoped rule carrying a source, a confidence tier and a before/after value — surfaced for review, where you accept, reject or edit it. Export is gated behind an explicit confirmation, so nothing leaves the tool until you sign off. I built it end to end: a Python / FastAPI engine and a React front end, backed by a passing test suite.

• Local-first — processed on your machine; no accounts, analytics or telemetry
• Source-backed completion from CrossRef, OpenAlex and Open Library
• Every change carries evidence, a confidence tier and a reversible decision
• Reproducible run manifest + CSV / JSON / HTML audit logs

GetitDone — a productivity app, conceived and shipped solo.

Friction in my own academic workflow became a shipped product. GetitDone helps organise and prioritise work using an Eisenhower matrix — and I built and published it on the Apple App Store while finishing my PhD.

I ran the full product cycle: defining the problem, designing the UI and UX, iterating on user feedback, completing external review, and releasing. Four months of teaching myself AI-assisted coding, turned into something real.

• End-to-end product ownership, from problem to release
• Swift & Xcode, version-controlled on GitHub
• AI-assisted development with Claude & ChatGPT
• Iterated UI / UX from real user feedback

Agent Mindset — learning AI for law, in the open.

Agent Mindset is my working notebook for teaching myself how AI actually works and how it applies to legal practice — published openly as I go. It pairs self-paced study modules (from how language models tokenize and reason, through to hands-on courses on legal-AI platforms like Harvey and Legora) with an automated daily digest of legal-AI developments and a transparent build log that documents the failures as well as the fixes.

It is also a live demonstration of responsible automation: Claude agents curate and update the daily digest, while the content stays structured as validated JSON and markdown so nothing drifts. The same communication instinct that took my research to national media, aimed squarely at the technology I am most curious about.

• Study modules on LLM internals — tokenization through inference
• A multi-part course on legal-AI platforms (Harvey, Legora)
• An automated, filterable daily digest across 25+ tracked sources
• A public build log — experiments, breakages and fixes in the open