Eluviant Debuts Aurora Flow AI Model as It Rebrands From IntelexVision
AI company Eluviant has unveiled Aurora Flow, a generative AI model built for enterprise-scale video surveillance, alongside a corporate rebrand that retires the IntelexVision name after nearly a decade in the market. The model is already running in live deployments and can process footage across multiple cameras in near real time, including in fully isolated environments with no external network connectivity.
Understanding events, not frames
Aurora Flow is designed to interpret sequences of movement and behavior rather than individual frames, an approach that Callum Wilson, founder and co-CEO of London-based Eluviant, told SecurityInfoWatch helps the system understand incidents that cannot be accurately assessed from a single image. Asked for an example of where that difference changes the outcome for a security operator, Wilson pointed to ATM tampering.
“With sophisticated fraudsters, the behavioral differences between what would appear to be a ‘normal’ transaction and tampering are subtle,” Wilson said. “But there are tells in their hand placement, posture and attention patterns over time that distinguish their actions from how a typical customer would interact with the machine.”
Rather than evaluating isolated images, Aurora Flow analyzes complete event sequences to determine whether activity represents routine behavior or an incident requiring attention, according to Wilson. That gives operators alerts with meaningful context and urgency instead of generic notifications that still require manual review, helping them prioritize critical events and make faster, more informed decisions. He described the same approach applying to scenarios such as store theft, aggressive behavior and dangerous driving, where understanding how an event develops matters more than what appears in any single frame.
Aurora Flow builds on Eluviant’s existing platform, which pairs an unsupervised self-learning engine capable of flagging genuinely unforeseen events with Aurora, the company’s vision-language model that has been part of live alert decision-making for the past 18 months. In a press release, Eluviant said the model can also help identify behaviors such as equipment tampering and unsafe climbing in hazardous environments. The company’s technology is currently used across sectors including retail, critical infrastructure, construction and smart cities, supporting functions such as loss prevention, anomaly detection and incident response.
A rebrand years in the making
The launch coincides with the retirement of the IntelexVision name. Founded in 2017, the company becomes Eluviant, a change Wilson said reflects its evolution beyond traditional video analytics. Asked why now was the right moment for the shift, he said the company spent the past eight years proving its technology in demanding environments, building customer and partner relationships and demonstrating that AI could deliver real value in surveillance.
Since then, its technology has expanded well past video analytics into a platform designed to help organizations turn existing camera networks into systems that learn, understand complex events, support real-time decisions and provide operational insights beyond traditional security use cases.
“Eluviant is an evolution, not a departure, and represents the transition we have made to become a global leader in enterprise-scale video intelligence,” Wilson said.
Adoption hurdles and North American growth
Widespread adoption of generative AI in physical security will depend on more than advances in the models themselves, Wilson said, pointing to two separate hurdles. The first is trust: organizations need confidence that a system will hold up across varied facilities, cameras and lighting conditions and integrate with existing infrastructure, since a system that adds noise, requires constant adjustment or operates in isolation is unlikely to earn its keep. The second, he said, is less about the technology and more about how organizations think about what their camera networks are for in the first place — a point he returned to later when describing his vision for how video surveillance should function inside the enterprise.
Eluviant is also expanding its presence in North America through a combination of channel partnerships and a regional sales organization. Wilson said the company appointed former Evolon Sales Director Bob Finocchioli earlier this year to lead its go-to-market efforts in the United States and Canada. The company’s priorities in the region, he said, are to strengthen and expand its partner network while increasing awareness of the broader potential for AI-enabled video intelligence within physical security.
Wilson believes organizations will need to rethink the role video surveillance plays within the enterprise.
“With video intelligence, it’s possible not just to know exactly what happened by the second across every site, but to apply your organizational context to that data so you can make sense of it and turn it into insight you can use across departments,” he said. “But tapping into this new well of data requires seeing cameras as operational tools and shifting to a more collaborative way of working with departments outside of security.”
About the Author
Rodney Bosch
Editor-in-Chief/SecurityInfoWatch.com
Rodney Bosch is the Editor-in-Chief of SecurityInfoWatch.com. He has covered the security industry since 2006 for multiple major security publications. Reach him at [email protected].