Turbine

Data Pipeline for LLM Apps

Elevate your business by using your data to power AI bots. Provide rich and up-to-date context to your LLM apps using our fully-managed data pipeline.

Watch a Demo of Turbine

Features

Turbine lets you create pipelines to sync data from any source to any vector database. Pipelines are fully configurable, and lets you bring your own data source, embedding model, and vector database.

Integrates seamlessly with your existing data sources

You can connect S3, PostgreSQL and MongoDB to Turbine currently, with more integrations coming soon.

Database changes synced instantly

Stop using crude batch jobs. Turbine uses state of the of data engineering pipelines that syncs changes to your database in real-time.

Bring your own embedding models and vector indexes

Turbine lets you use your own embedding models and vector indexes. We support Pinecone, Milvus, OpenAI and HuggingFace right now, with more coming soon.

Endless configurability

You can tweak and optimize Turbine according to your use case: filters to apply over incoming data, fields to include in the embedding, chunking strategy to use—everything is configurable.

Blazing fast and designed to scale

Turbine is designed from the ground up to be scalable, using modern distribute stream-processing platforms to handle all moving data.

Start easily with Turbine Console or HTTP API

You can get up and running with Turbine under 2 mins with a single HTTP POST request. Or you can use Turbine Console's intuitive UI if you prefer.

The Turbine API is as simple as this

Turbine code sample for Python

Head over to Turbine console to get started.