We just launched Turbine, it automates the data pipeline for LLM powered apps. It fetches data from your database, creates embeddings from the data, and stores in a vector database for easy semantic search. It also creates a real-time data pipeline to fetch changes and keep the search data fresh.

Turbine supports multiple source databases, embedding models and vector databases. It's aimed to be configurable and easy to use at the same time.

It's primary use case would be being the data backend for LLM apps—to create a relevant context for each prompt from your data.

We are very early and this is our first public launch, we just have tested the product internally. Any and all feedback would be very appreciated. DM or email me at sumit@useturbine.com if you are interested to be a early beta user of Turbine and help us build this product.

Thank you! Looking forward to your comments.