V7 Labs, the company behind a computer vision platform designed to streamline and future-proof the training data processes for artificial intelligence teams as AI technology evolves, has secured $3 million in funding. This seed round was led by Amadeus Capital Partners, with additional investment from Partech, Air Street Capital led by Nathan Benaich, and Miele Venture.

Established in 2018 by Alberto Rizzoli, a graduate of Singularity University, and Simon Edwardsson, formerly the R&D lead at RSI (the creators of the “seeing” application Aipoly), the V7 Labs platform aims to accelerate the creation of superior training data by a factor of 10 to 100. It achieves this by allowing users to construct automated pipelines for image and video data, effectively manage and version intricate data sets, and both train and implement cutting-edge vision AI models.

Rizzoli of V7 Labs explains that continuous data collection, labeling, and model retraining are essential for businesses developing computer vision solutions that generate tangible results. He notes that when Aipoly was developed in 2015, the team was compelled to create their own tools and stay current with the latest AI advancements due to the lack of available third-party SaaS solutions.

Rizzoli observes that many leading computer vision companies are now utilizing SaaS platforms like V7 to address this challenge. He points out that startups focused on AI have numerous considerations, and efficiently storing and accessing large volumes of video data is a concern that typically arises during the service delivery phase.

“V7 encapsulates established industry standards for data organization, labeling, and the deployment of computer vision models to tackle practical, real-world issues.”

“Users will soon have the ability to audit these models – and their associated training sets – to identify and resolve issues, assess data quality, pinpoint failure scenarios, and mitigate any potential biases,” he adds, highlighting these as significant unresolved challenges within the AI sector.

Currently, V7 Labs serves over 100 customers, including Tractable, GE Healthcare, and Merck. The platform is experiencing the most rapid growth in the medical imaging field, partly due to its support for DICOM annotation and HIPAA compliance, both critical requirements in healthcare.

However, based on the volume of data processed, Rizzoli indicates that “expert inspections” represent the most prevalent use case. “This includes numerous companies employing AI to detect damage or irregularities in vehicles, oil platforms, power lines, pipelines, or roadways,” he explains.