I'm a full-stack engineer in Berlin. I've spent five years building web products, TypeScript and React on the front, Node and Python behind, and the last two building and evaluating LLM agents in production. I'm interested in the unglamorous parts: what happens when the model is wrong, how you know, and what you promise a user you'll never do. I built Vera Civic, a legal assistant for Nigerians, and attest, an open-source evaluator that grades an agent against what its tools actually returned.

Currently at Creator.co, building Vera Civic, and maintaining attest.

Projects VeraCivic A legal and civic assistant for Nigerians: ask in English, Pidgin, Yoruba, Hausa or Igbo and get the Act and Section that applies, word for word.AI Influencer Matching at Creator.co The creator-matching page in Creator.co's brand dashboard, where London, the platform's AI agent, returns ranked creators. I built the frontend and the server pieces it needed, on a global team.attest An open-source library and CLI that grades an AI agent's run against the evidence it recorded, claim by claim, not on how good the answer sounds.Discord Music Agent A Discord music bot that turns "some chill vibes", a half-remembered lyric or a Spotify link into a queue of studio masters, played bit-exact and gapless.HerPro Marketplace Platform The backend for a Nigerian marketplace and food-delivery app: customers, vendors and riders, Paystack payments, KYC, wallets and push notifications. Node and TypeScript on MongoDB.Emotion Detection for Autism Support Three CNNs trained from scratch on FER2013, CK+ and AffectNet, wrapped in a web app that classifies faces from a photo, webcam or video.
Writing Aug 2026 Evals as Tests, Part 2: What Happens When the Prompt Injection Is Inside a PDF?Jul 2026 Beyond Vector Similarity: Ranking Authority, Currency, and Relevance in Legal RAGJun 2026 Evals as Tests: Grading Your Agent with attest All writing →
Elsewhere
© 2026 Adepeju Orefejo