We build, break, and ship AI products in public.

Website  ·  X  ·  Discussions  ·  Contact

nexlabz is a product lab. We start with a question — what if this actually worked? — and turn it into something you can use, usually in weeks rather than quarters.

Most of what lives here is an experiment. Some of it graduates into a real product. Some of it gets deleted, and we say why. Both outcomes are the point.

_{🌱 Idea  ·  🧪 Experiment  ·  🚧 Building  ·  🚀 Live  ·  🪦 Archived}

1. Ship the ugly version first. A rough thing in the world beats a polished thing in a doc.
2. Public by default. Repos, decisions, and postmortems are open unless there's a reason.
3. Small surface, deep value. One thing done properly instead of five done adequately.
4. The model is a component, not the product. Good AI products are mostly good products.
5. Kill things early. An experiment that ends cleanly is a success, not a failure.

Everything public here is open source and free to fork, read, and learn from. Check each repo's LICENSE — most are MIT, a few differ.

If you're building on something of ours and it breaks, open an issue. If you want to change it, read CONTRIBUTING.md first — it's short.

We're a small team and we like working with people who ship. If you've built something we'd find interesting, or you want to collaborate on an experiment, reach out at hello@nexlabz.dev.

Found a security issue? Please don't open a public issue — see our security policy.

_{Built by nexlabz. Mostly at night.}