Building with LLMs and tool use
Hands-on experimentation with how large language models make decisions, call tools, and compose answers from real data sources.
What I'm interested in
Areas I'm actively exploring
Connecting systems and data
Using public APIs to build things that fetch, transform, and display live information — no mock data, no simulations.
Shipping things in plain HTML, CSS, and JS
Building projects without frameworks or build tools so the mechanics are always visible and nothing is hidden behind abstractions.
Coming up
What's in progress
More AI agent experiments
Expanding the MCP pattern to other APIs — more data sources, more interesting questions, more complex tool use flows.
Project writeups
Detailed walkthroughs of how each project was built, what broke, and what was worth learning from the process.
New tools and stacks
Trying out technologies I haven't worked with before — each project is a chance to pick up something new by actually shipping it.
Say hello
Have an idea or just want to connect?
I'm always happy to talk about what I'm building, swap notes on AI tooling, or just chat.