Ship reliable LLM agents and applications.

We provide a single platform to build, deploy, monitor, and evaluate complex LLM agents.

Key features

  • Versatile pipeline builder: Build LLM pipeline using variety of nodes with required functionality provided out of the box.
  • Semantic Search: Cache data or upload it, so that it can be quickly accessed based on semantic similarity. Fully hosted by us.
  • Evaluations: Evaluate your LLM pipeline with custom evaluator pipelines.
  • Datasets: Store your data for use in evaluations or for semantic retrieval.
  • Endpoints: Deploy your pipeline to an API endpoint in seconds to our scalable inherently fast backend.
  • Logging: Conveniently access logs and see how users interact with your pipeline through endpoints.

To see all features, you can read more in other sections.

Getting started

Read our getting started tutorial to onboard.

We are adding new features every day. Feel free to contact us as or ask us in our discord community for any feature requests.