AI company Eluviant has launched Aurora Flow, a generative AI model designed for enterprise-scale video surveillance deployments, saying the system can operate across multiple cameras in near real time, including in isolated environments without external network connectivity.
In a media release, the London-based company said Aurora Flow is intended to address a common limitation in video analytics: interpreting events across sequences of movement over time, rather than identifying objects or actions in individual frames.
Eluviant said the model builds on its existing platform, combining an unsupervised self-learning engine intended to identify unforeseen events with its vision-language model, Aurora, which it said has been used in live alert decision-making for the past 18 months.
The company said potential use cases include detecting equipment tampering, unsafe climbing and dangerous driving, particularly in “secure and sensitive environments”.
“We believe Aurora Flow is a frontier AI model in surveillance and a step change in what video intelligence can deliver, moving beyond detection and into genuine understanding of behaviors and actions in complex real-world environments,” Callum Wilson, founder and co-CEO of Eluviant, said. “It addresses a challenge that traditional video analytics has struggled to solve efficiently: understanding what is happening when a single still frame is not enough.”
Eluviant also used the announcement to confirm a rebrand from IntelexVision, stating the new name reflects changes in video intelligence driven by AI and the company’s future direction. “Everything we have built under the IntelexVision name — our technology, customer relationships, and hard-won understanding of what enterprise deployments actually require — is the foundation this moment stands on,” Wilson said.
The release said enterprise adoption is being driven by organisations seeking to extract operational value from existing CCTV deployments beyond security, and cited a forecast that the market for enterprise-scale video intelligence could reach $30 billion by the end of the decade.
