Introducing Vpuna AI Search: Revolutionizing Semantic Search!

A long-time fan of the HN community, the author has launched a developer-oriented semantic search platform that allows users to upload and index both structured and unstructured documents. Key features include managing tenants and projects, uploading various file types (with more formats coming soon), and accessing three APIs for search, document upload (with embeddings), and deletion. The platform uses CPU-based sentence-transformers for embeddings, with future support for more models and features like LLM summarization. The motivation behind the project stems from client needs for advanced semantic search beyond basic keyword functionality, highlighting existing tools’ limitations in cost and flexibility. The author seeks feedback on the initial release and suggestions for improvements, inviting interested users to explore the README linked in the post for more details.

Source link