A billion cameras, and no one really watching

There are now more than a billion surveillance cameras installed worldwide, and that number keeps climbing. Every one of them generates a video feed that, in theory, someone is watching for the moment something goes wrong. In practice, a single control room operator can only hold useful attention on a live video feed for a matter of minutes before performance drops off, and the traditional response, adding more cameras and more screens, only makes the underlying problem worse. Most video analytics tools on the market compound this: built around deep learning models trained for one narrow, predefined use case, they are computationally heavy, expensive to scale, and blind to anything that does not match a pattern they were trained on.

Filtering the noise before asking AI to look at it

Eluviant, founded in the UK as InteleXVision, takes a different approach to the same problem. Rather than running every frame of every camera through a deep learning model, the company's platform, Sentry, uses a self-learning layer that establishes what “normal” looks like for a given camera's field of view without any manual configuration, then filters out the vast majority of the video stream before it ever reaches a more expensive analysis step. Only the genuine anomalies, footage that deviates from the learned baseline, get passed on for deeper contextual analysis and flagged to a human operator. That two-layer design lets a single operator effectively cover far more cameras than a traditional control room, and it works across a wide range of environments, from perimeter and area protection to wide-area surveillance, without requiring the heavy rule-based pre-configuration that older analytics tools demand. More recently, the company has added Aurora Flow, a generative AI layer built to understand events and behaviour as they unfold over time rather than evaluating isolated frames, extending the platform from anomaly detection toward genuine situational understanding.

From InteleXVision to Eluviant

Founded in 2017 by Callum Wilson and Michael Vorstman, the company builds a video intelligence platform designed to operate at the scale the surveillance camera market actually requires, combining deep technical expertise in AI and computer vision with hands-on experience in the video surveillance industry. Run from London, the company is a genuinely international operation, deploying its technology across multiple continents and industries including transport, energy, manufacturing, logistics and healthcare.

A visionary founding team supported by experienced operators

Eluviant is led by its co-founders, Callum Wilson and Michael Vorstman, who have built the business from its early video-analytics work into a video-intelligence platform deployed across complex enterprise environments. They are joined by Graham McGregor, previously CFO at AllPoints Fibre, as Chief Financial and Operating Officer, strengthening the company's financial and operational leadership, and by a senior commercial team with deep market knowledge: Alberto Croso, who has held C-level positions at world leading security integrators, and Rafik Lamri, who leads the META region and brings decades of experience selling security software. Together, the team combines the founders' long-standing product and customer understanding with operational discipline and international go-to-market experience.

When we invested

We backed the company in 2021, in its early funding round in the UK alongside Inveready, drawn to the strength of the founding team and the technical differentiation of Sentry's self-learning approach at a time when most competing products were locked into rigid, rule-based analytics. The company has continued to raise capital since, including a Series A in early 2025 that brought in Acurio Ventures alongside Adara Ventures and Inveready, as it pushed into new regions and industries.

Where things stand

We backed a bet on smarter, self-learning video analytics, and the company has built that into a broader video intelligence platform extending into generative AI with Aurora Flow. The thesis holds: physical security infrastructure, cameras above all, is exactly the kind of real-world system that benefits from AI applied thoughtfully rather than reflexively.