The hub for your machine learning stack

Aquarium integrates seamlessly with your machine learning model pipelines. Monitor and orchestrate your ML data operations all from one place.

Automate manual workflows with code

  • Python client API: Automatically upload and access datasets, model inferences, metadata, embeddings, segments, and more.
  • Made for developers: Get started with Aquarium in less than a day. No need to copy unstructured data from place to place.
Check out our docs page

Explore our Docs.
Learn how we can help you and your systems.

Work hand in hand with labeling

Webhooks and API integrations: Click a button in the Aquarium UI to export data to your labeling system.
Streaming dataset uploads: upload new datapoints into a versioned view of your dataset that retains the history of previous changes.

Freedom to choose

  • Use your favorite machine learning tools with Aquarium’s integrations and flexible API.
  • Evaluate different vendors and home-grown systems. If you find a new favorite, it's easy to switch.

All in one place

  • Have a comprehensive view of your ML development cycle. Aquarium helps you understand what to do next and enables you to take actions to improve your ML system.
Collaborate with your ML team in a single place. No need to back and forth between task trackers, notebooks, and spreadsheets.

Get in touch

Schedule time to get started with Aquarium