Mini LLM Educational Kit

A lightweight, pre-built LLM (~9M parameters) that trains in minutes on free cloud hardware, designed to teach people how language models actually work. Solves the complexity barrier for AI learning. Target users are students, hobbyists, and educators wanting hands-on ML experience.

Demand for hands-on AI education is real and accelerating — courses like fast.ai and Andrej Karpathy's makemore/nanoGPT tutorials have proven that people want to *build* small models, not just use large ones, and that appetite has only grown post-ChatGPT. Karpathy's nanoGPT is the closest substitute and it's free, well-documented, and already beloved by the exact audience being targeted here, which is the central problem. The revenue band is unknown for a reason: educators and hobbyists are notoriously resistant to paying for tooling they expect to be open-source, and the realistic monetization paths — cohort courses, a paid tier, institutional licensing — each require a different go-to-market entirely. The most likely failure mode is building a slightly friendlier wrapper around something people are already happy to get for free, without a clear reason to pay.