SemaDB

Hello Product Hunt Community 👋 We’re excited to share SemaDB version 2 and it’s now fully open source! It is designed to handle complex vector queries that go beyond basic hybrid search such as: “Find me the top 5 nearby restaurants that are open now, have a rating of 4 or more, and serve a dish similar to this image or have a similar description to the user query within the category Chinese or Japanese weighted by image then description and sorted by price.” Combine vector search, filtered search, full text search and more into a single query to leverage the most out of our flexible document based indexing. It is extensible with more index types and features ranging from geo indexing to automatic sharding and true multi-tenancy support. With a commitment to open-source, we hope it can serve as a building block to many AI applications and as a learning tool for new developers interested in vector databases. You can get started with a single command to run locally or try out the fully hosted SemaDB Cloud currently in beta ☁