Document Q&A Tool
A retrieval-augmented Q&A tool: ask questions about your own documents and get answers grounded in them, with citations. Local embeddings, conversation memory, sentence-aware chunking, plus a CLI and a web interface.
AI Consultant in training
I'm 23 and learning how to implement AI inside real businesses, with the goal of launching my own consulting firm doing exactly that. Eager to take on new projects and stack up knowledge along the way.
Live systems
Full status page →I'm Liam — a last-semester business student at Stockholm Business School, studying International Business and Politics, graduating in January 2027. Outside coursework, I'm focused on one thing: figuring out how AI actually creates value inside a business, not just in theory.
That means testing tools on real problems and building the kind of hands-on experience a client can trust — the tools on this page are the proof, not just the plan. Long-term, the goal is my own AI consulting practice: understand the real problem first, then build something that survives contact with it.
Real projects, each a different technique — retrieval, classification, generation, and more as they ship.
A retrieval-augmented Q&A tool: ask questions about your own documents and get answers grounded in them, with citations. Local embeddings, conversation memory, sentence-aware chunking, plus a CLI and a web interface.
A small Claude-powered service that reads an inbound message and classifies it — category, urgency, one-line summary — before it hits an inbox. Structured extraction rather than search: it's what's silently tagging the message you just sent through the form below.
Paste in a blog post, transcript, or article — get back a summary, 3 distinct social posts, and an email blurb, one Claude call. First project in the set that generates new content instead of reading or classifying it.
What actually goes into building and shipping this stuff, one project at a time.
Open to conversations, projects, and opportunities to learn.
Or reach me directly: