Show HN: I built a chat client that uses embeddings to cluster messages by topic

Category: ai-ml

Tags: chat-client, embeddings, clustering, llm, vector-similarity

Score: 4.7/10 (Innovation: 5, Technical: 5, Documentation: 3, Utility: 4)

A chat client that clusters messages by topic using LLM-generated summaries and vector embeddings to group related queries and responses. It uses cosine similarity against existing groups with a threshold-based assignment algorithm to organize conversation history into a navigable sidebar.

Target audience: AI/LLM power users, developers, researchers

Repository: https://llmtrail.com/demo

View on Hacker News