Qvyon launches with $8 million to test physical AI safely

Qvyon Technologies Inc. is entering the market for physical AI validation with a specific pitch: companies need to test smart machines against strange and dangerous conditions that are hard to reproduce in the field. The company's co-founder and Chief Executive Frantz Saintellemy said he spent years building AI for vehicles and saw how difficult it was to prove those systems were ready for the road. Mimic Core addresses this by ingesting camera, lidar and radar data to reconstruct environments in 3D and time-sequence snapshots, then generating challenging test conditions. This allows development teams to catch problems before putting robots to work in the real world.

The $8 million in funding, led by MaC Venture Capital with participation from BDC Capital, BKR Capital and Spotlight VP, will support Mimic Core and its rollout through the United States, Europe and Japan. Qvyon offers the product as software-as-a-service and as a self-hosted solution, and it stressed that customer data in its cloud product is kept separate and never used to train AI models for other customers. Beyond Mimic Core, Qvyon is developing Mimic Validation OS for the first half of 2027 to help companies measure and document test results, and Mimic Learn for late 2027 to help improve AI using information from machines already in operation. These planned products suggest Qvyon aims to cover the full validation workflow, from scenario generation to documentation and post-deployment learning. However, the company has not disclosed customer numbers or revenue, so its traction remains unclear.