Zumo Labs

We generate synthetic data for computer vision models.

Y CombinatorWinter 2020InactiveB2BEngineering, Product and DesignArtificial IntelligenceMachine LearningData Engineering

What they do

Problem

Computer vision models need large, diverse labeled training datasets that are costly, slow, biased, and privacy-sensitive to collect manually.

Solution

Zumo Labs generates feature-rich synthetic computer vision training data in the cloud to rapidly produce scalable, privacy-safe datasets including edge cases.

Who it's for

Companies and teams building computer vision systems such as autonomous vehicles, robotics, retail analytics, security, and other vision-based products.