Qvyon Raises $8M to Test Physical AI in Safety

Qvyon secures $8 million to accelerate the use of its simulation testing platform for robots and autonomous vehicles. The service is now available, with deployment announced in the United States, Europe, and Japan, along with two complementary products expected in 2027.

Accessible Service, Segregated Data, and Announced Deployment

Mimic Core is now offered via Qvyon Cloud, featuring a guided 14-day evaluation, and can also be installed on companies' internal systems. The product is available as software as a service and in a self-hosted version. Qvyon states that in its cloud, customer data is strictly segregated and is never reused for training models intended for other clients. The funding will support Mimic Core and its deployment across three regions: the United States, Europe, and Japan. The roadmap also includes Mimic Validation OS, scheduled for the first half of 2027 to measure and document testing results, and Mimic Learn, expected in late 2027 to leverage insights from machines already in production.

Simulation from Real Sensors to Multiply Scenarios

Mimic Core ingests recordings from cameras, lidar, and radar, then reconstructs 3D environments and chronological snapshots to replay scenes and their evolution. From this data, the platform generates challenging conditions to rigorously test AI models as closely as possible to the target hardware, producing traceable datasets and validation evidence. The system integrates with existing virtual testing tools and allows for testing models and configurations in a virtual environment without the need for costly field data collection. A single recording can be varied into numerous scenarios to identify failure points, helping teams spot issues before any real deployment.

Need for Validation and Funding Round Led by MaC Venture Capital

Qvyon has secured $8 million in funding aimed at assisting companies in the validation of robots and autonomous vehicles, during a funding round led by MaC Venture Capital, with participation from BDC Capital, BKR Capital, and Spotlight VP. The company, an expert in validating infrastructure for physical AI, targets applications in real-world conditions, such as factories, store aisles, kitchens, or roads. Before any deployment, understanding the behavior of systems under normal and abnormal conditions is essential, especially since reproducing rare and dangerous scenarios in the field remains challenging, such as a car appearing in front of a truck during a storm or a sensor stuck on a production line. Frantz Saintellemy highlights the difficulty, stemming from years of AI development for vehicles, in proving the ability to drive and emphasizes tests conducted with recordings specific to the vehicles and factories of the companies.