Data curation product for robotics that applies rigorous, reproducible quality analysis to select high-quality training samples.
Pareto is Hebbian Robotics' product for high-quality data curation in robotics datasets, addressing the problem that large collected datasets often go unused due to vague quality definitions and manual heuristic-based curation. It brings formal, reproducible analysis to robotics data quality, treating data study with the same methodology applied to model development. The company is backed by Y Combinator and investors from Oracle and Google DeepMind, with team members from Columbia, NUS, Stanford, Jane Street, and Verkada. Pareto is a product of Hebbian Robotics.