Bifrost builds a 3D data generation platform that creates simulated 3D worlds to train computer vision and robotics perception models, enabling teams to adapt models to new objects, tasks and surroundings faster. The platform is designed so AI engineers can generate training data without a dedicated 3D team, positioning it as an alternative to tools that require specialized 3D operators. Bifrost is currently in closed beta with select heavy-industry partners and plans a public launch in the coming months while hiring to speed product development. Its primary users are AI developers focused on robotics, computer vision and perception across sectors including robotics, aerospace, defense, maritime, geospatial and industrial automation; target customers include large industrial companies, government organizations and growth-to-late-stage startups. The company sells its platform via an annual subscription model and has 22 staff across the U.S. and Singapore. Founders Charles Wong and Aravind Kandiah started Bifrost in 2020; Wong previously worked on perception at Nutonomy and Kandiah built a medical AI system for detecting diabetic retinopathy.

Total raised

$8M

Funding rounds

1

Latest round

Series A

Latest activity

Oct 2024