Improve your datasets without having to write code. Aquarium’s web user interface empowers your data operations team to curate datasets without help from engineering.
Free up your ML engineers to improve infrastructure and develop improvements to the model code.
“With Aquarium, we are more confident that we’re consistently delivering good models."
Dan Grollman, Principal Engineer at Plus One Robotics
Set up ML regression tests
Catch problematic outputs before shipping to production. Ensure that each new model iteration is better than the last.
Never regress performance on important cases as you iterate on your ML system.
Version control for data and models
View previous snapshots of your dataset and see diffs between versions. Aquarium stores the history of changes to your datasets and models.
Track trends in model performance over time. Figure out which initiatives are working and which are not.