ai-knowledge-graph

Why we included this project

This project turns an unstructured text file into an interactive knowledge graph without asking you to maintain a full graph database pipeline. You point it at a document, pick any LLM that speaks the OpenAI API format, and it extracts subject-predicate-object triples and normalizes entity names across chunks. Beyond that, it can infer connections that weren't spelled out in the source. The output is a single HTML file you can open in a browser, and the README links a working demo so you can see what the result looks like before committing. Because it works with local models through Ollama or vLLM, it's a reasonable choice when the source material is sensitive or private and you want to keep processing on your own hardware.