Enlighten-AI is a personal, experimental project where I built a retrieval-augmented generation (RAG) assistant with a supportive, coaching-style voice. It’s for anyone curious about how RAG can ground a language model’s responses in a curated body of source material rather than letting it improvise freely.

The assistant retrieves from a curated set of Dr. K (HealthyGamer) transcripts and uses them as context for generation, so answers stay anchored to that material. The repository is mainly a sandbox for me to explore the practical side of RAG: chunking, retrieval, and prompting a model to respond in a consistent, empathetic tone. It is not a medical or therapeutic tool, and it isn’t meant for production use.

This one sits a bit apart from my core visual-attention research, but it reflects my broader interest in large language models and how retrieval can make them more reliable and context-aware. Consider it a work-in-progress exploration rather than a finished product.

Explore the code on GitHub.