Dr. Timnit Gebru speaks during a talk titled “The Human Cost of Smarter Systems” at AfroTech 2026

Timnit Gebru Talks Power, Bias, and the Human Cost of AI at AfroTech

November 13, 2026
Dr. Timnit Gebru speaks during a talk titled “The Human Cost of Smarter Systems” at AfroTech 2026

Timnit Gebru Talks Power, Bias, and the Human Cost of AI at AfroTech

November 13, 2026

Qvyon Technologies Raises $8 Million to Validate Physical AI Systems

NewsWorld
Qvyon Technologies co-founder, Frantz Saintellemy's headshot

Qvyon Technologies co-founder, Frantz Saintellemy's headshot. Image Credit: Shoppe Black

Montreal-based Qvyon Technologies has raised US$8 million in seed financing to advance MIMIC, its platform for testing and validating physical AI systems such as robots and autonomous vehicles.

MaC Venture Capital led the US$6 million equity portion, with participation from BDC Capital’s Black Entrepreneurs Fund, BKR Capital, and Spotlight VP. RBCx provided the remaining US$2 million in venture debt.

The startup, which has emerged from stealth, plans to use the financing to expand into the U.S., Europe, and Japan and develop new products for its MIMIC platform.

Related Post: British-Canadian OXcan Secures $11 Million For Lung Cancer Detector

The Problem of Validating Physical AI Before Launch

Physical AI refers to intelligence that extends beyond the digital space and into the physical world. It enables smart machines to sense, think, and act in real-world environments such as factory floors, retail aisles, kitchens, and roadways. 

Before companies can deploy smart machines like robots or self-driving cars in the real world, they need to understand how these systems will behave in both normal and unusual situations. Simulating rare, unexpected, and dangerous conditions in the field is difficult and risky.

Advertisement

For instance, a car could suddenly cut in front of a truck during a rainstorm, or a robot sensor array could become blocked or damaged while working on a production line. To address this, Qvyon introduced Mimic Core, a platform that turns real-world sensor recordings into simulated environments and generates varied test scenarios. This lets developers and engineers access data and test physical AI models and robotics setups in virtual settings without repeatedly collecting field data.

“We spent years building AI for vehicles and saw how hard it was to prove those systems were ready for the road,” said co-founder and Chief Executive Frantz Saintellemy in a press release. “Companies need to test what happens in their own vehicles and factories by using their own recordings.”

Related Post: BKR Capital Closes $20M for Black Innovation Fund II

Advertisement

How Qvyon’s Platform Works

With just one recording, an engineer can create many test variations and pinpoint exactly where a system makes mistakes. This helps development teams catch problems early, before putting robots to work in the real world.

Qvyon system takes in camera, lidar, and radar data, then rebuilds environments in 3D and in time-sequence snapshots that capture both the scene and how it changes over time. It then uses those reconstructions to create difficult conditions for testing AI models against the hardware they will control in simulation.

The platform also produces traceable datasets and validation evidence that engineers can use to judge how smart machines perform before deployment. Mimic Core fits alongside the virtual testing tools companies already use, helping them better understand the scenarios and conditions robots and smart machines will face.

The Montreal-based company offers the product as software-as-a-service and as a self-hosted option for companies that prefer to keep everything in-house. Qvyon emphasized that with its cloud product, customer data stays separate and is never used to train AI models for other customers.

Early customers include NuPort Robotics and Voltava, and Qvyon plans to expand commercially across the U.S., Europe, and Japan while developing MIMIC Validation OS (H1 2027) and MIMIC Learn (late 2027).

Advertisement

Related Post: Meet the Enterprising Owner-Operator Behind Ottawa’s First Chick-fil-A’s Restaurant

Next Steps for Qvyon Technologies

“Platform shifts get decided by the infrastructure underneath them, not by the models on top,” Marlon Nichols, Co-Founder and Managing General Partner at MaC Venture Capital, said.

“Every company deploying AI into the physical world is about to hit the same wall: proving the thing works before it ships. Frantz and his team built what they wished they’d had when they were the ones trying to prove it. Whichever perception model wins, it will need what Qvyon is building.” 

According to the company, the funding will support Mimic Core and its expansion across the United States, Europe, and Japan. Qvyon is also working on Mimic Validation OS, expected in the first half of 2027, to help companies measure and document test results, along with Mimic Learn, expected in late 2027, to help improve AI using information from machines already in operation. Mimic Core is available now through Qvyon Cloud with a 14-day guided evaluation, or for companies to run on their own systems.

Advertisement

Join the UrbanGeekz community for the latest in tech, innovation, entrepreneurship, culture, and opportunities. Subscribe here.

Stephen Oluwadara
Stephen Oluwadara
Stephen Oluwadara is a general news reporter for UrbanGeekz covering stories across the US and Africa.
Share
Share
Tweet
Reddit
Email