I build multi-agent AI systems things that research, reason, and produce outputs without hand-holding. Mostly Python, LangGraph, and whatever API is most annoying to wrangle.
MSc Data Science @ University of Westminster. Based in London. Currently shipping agentic AI at Cyber Scallywags.
What I've been working on:
- Multi-Agent Research Assistant — four LangGraph agents (Supervisor, Researcher, Writer, Critiquer) that collaborate to research a topic and self-review the output until it's good enough
- Startup Idea Validator — describe your idea, agents check market data via Firecrawl, HN sentiment, and GitHub feasibility, then give you a structured verdict
- Course Platform — full-stack: FastAPI backend, Next.js frontend, Stripe payments, JWT auth, PostgreSQL
- AI Revision Platform — upload your notes, get quizzes and summaries back


