Eleven entries, chosen and described by hand. Each description is condensed from the project's own README, and public repositories link straight to the source.
Turns a scientific hypothesis into a structured, operationally grounded experiment plan. Took second place at the MIT Club of Northern California and MIT Club of Germany hackathon, Challenge 04.
An AI-powered social media content agent, entered in the same hackathon; it did not place.
A Cursor ruleset and workspace scaffolding for spec-driven, AI-assisted development: one workflow from constitution and PRD through tasks to verification, with quality gates.
A project with two goals held together: studying scripture systematically, and learning to build agentic AI systems through real work rather than tutorials.
An exploration of a decentralised, AI-assisted approach to identifying misinformation.
A hands-on AI training course for business professionals with no coding background, covering agents and retrieval-augmented generation. Built on TypeScript, React and Firebase.
An open-source personal AI assistant. Not my project: I contributed 55 commits to it between March and April 2026.
A Home Assistant integration combining NASA public APIs, satellite tracking and sky visibility calculations into a glanceable view of what is overhead.
A Home Assistant integration that surfaces an Easy Equities and Satrix portfolio on a personal dashboard.
An offline-first Android budgeting app built around envelope budgeting and debt snowball methods, with cloud sync.
A Streamlit application that converts scanned weight certificate PDFs into structured JSON using OCR, with optional schema validation.
Practical systems thinking for leaders and builders navigating AI-driven change.
Most of my work lives in private repositories, so the public projects above are a partial view. Client and commercial work is not shown.