Vizgard is the sort of company ministers like to point to when they talk about commercialising defence tech, but the business case matters more than the slogan. The London SME, founded by former Royal Navy submariner Alex Kehoe, has spent the past five years building software aimed at one of the least glamorous problems in drone operations: how to help overstretched operators process more visual data, more safely, and with fewer missed threats. According to the UK government case study, UK Defence Innovation funding helped move that work from a promising idea into a usable product. For Market Pulse UK readers, the more interesting point is not simply that public money was awarded, but that it arrived at the stage when an early defence software firm needed proof, references and time to build.

Vizgard's platform, FortifAI, is designed to sit on top of existing camera infrastructure across land, air and maritime settings. In practical terms, it uses visual AI to automate detection, tracking and threat identification, reducing the amount of constant screen-watching expected from human operators. That matters because the commercial opportunity in drone tech is no longer just about hardware. The value is increasingly in software that makes fleets safer, more scalable and easier to run beyond visual line of sight. In the source article, the initial use case was supporting police drone operations, but the same operational problem appears across defence, security and critical infrastructure.

The turning point came through the UKDI Security Open Call. Kehoe said in the government case study that the company had to rebuild the software architecture so FortifAI could move from handling a single camera stream for a single drone to managing multiple streams at once. That sounds technical, but the business significance is straightforward. A product that works in one demonstration is not yet a market proposition. A product that can handle dozens of concurrent streams, and now more than 100, starts to look like something a large customer can deploy at scale.

Earlier backing had already given Vizgard an unusually valuable asset for a young defence-tech firm: a first commercial contract. In 2021, UKDI supported sea trials on the Royal Navy's uncrewed test submarine, where Vizgard combined radio tracks with visual AI to identify vessels obscuring their identity and to follow fast-moving craft automatically. For investors and procurement teams, that sort of field validation tends to matter more than a polished pitch deck. It gave the company a live reference point, and Kehoe said it helped open conversations with larger prime contractors working in coastal security.

The operating case for FortifAI is easy to grasp. Drone and counter-drone teams are often inundated with live feeds, while a human observer can miss an object simply by looking at the wrong screen at the wrong moment. Vizgard's software is meant to act as a triage layer, flagging where attention is needed rather than asking staff to watch everything, all the time. If that promise holds in wider deployment, the return is not only tactical. It also speaks to labour productivity, training time and the cost of running more complex missions with relatively small teams. That is often where software businesses in the defence supply chain begin to separate themselves from one-off project work.

The growth story around the company suggests that public R&D support did more than fund experimentation. The government case study says Vizgard grew from roughly five employees when it first worked with UKDI to 20 staff, a meaningful change for a specialist SME operating in a talent-hungry part of the market. Private capital followed. In early 2025, Vizgard secured £1.5 million in venture funding to build UnifAI, a machine-learning operations platform intended to cut the time needed to improve AI models in the field from weeks to days, including for users without deep technical expertise. That is a useful signpost: once the initial product was de-risked, the company could raise money for the tooling and workflow layer around it.

By 2026, the business had moved beyond a purely domestic story. Vizgard was exporting through a US Defense Innovation Unit contract for field testing with the US Marines, while also securing a place in NATO DIANA's 2026 cohort for autonomous and unmanned technologies. The programme accepted 150 companies from more than 3,500 applicants, giving the firm another credibility marker as it tries to grow internationally. Closer to home, UKDI support also led to work with the Defence Science and Technology Laboratory on edge-based AI for quadcopters identifying unexploded mines during automated intelligence, surveillance and reconnaissance missions. Kehoe's broader point in the source article was simple enough: for an SME, fully funded projects tied to real end-user problems can speed up research, sharpen the product and put founders in front of customers sooner. Vizgard's progress suggests that, in this case, the model has done exactly that.