NLT·01 — Live RAG demo

Lodestar

A live AI assistant grounded in this site’s content. Named for the star sailors once navigated by, it demonstrates retrieval-backed answers in a working product.

RAG Self-hosted embeddings pgvector Multi-provider LLM

Answers are grounded in this site’s own content via retrieval — not general web knowledge. If Lodestar doesn’t know something, it’ll say so and point you to the contact form instead of guessing.

This is a quick, fixed demo of the concept. Want to see it work on your own documents? Try the full Lodestar Workspaces demo →

Lodestar — How this works

Before answering, Lodestar searches this site’s source content for relevant material and generates a response grounded in what it retrieves. Search runs on self-hosted infrastructure, while the final response uses a hosted AI model with automatic provider fallback for resilience. The application operates as an independent service alongside the main site.