01732 759725 magazine 25 AI He adds that half of ControlUp’s 400 employees are inside the company’s product development group and that that engine is ‘running really hot’. “We’re building smaller teams that are activating much faster, and we have customers that ask for something on a Friday and by Monday it’s in the product. That’s the type of innovation speed that you can expect to see from us going forward.” Notable launches include ControlUp AI Assistant, an LLM chatbot for the data within Controlup that can also connect to AI MCP servers and interrogate other systems. The first feature powered by ControlUp’s Pulse AI agentic intelligence layer, it allows IT teams to ask AI, in plain English, for information on what is happening within the IT estate including advice on how to fix things. AI is also powering a new self-service feature, Controlup Connect, that for the first time links an employee to ControlUp so that they can fix certain issues themselves. AI use cases Ayres’ priority now is to generate durable use cases that show how AI is helping customers to speed up remediation and save money. “That’s our focus right now, to walk into every organisation with these agents laid down collecting data and showing a path to fewer tools, fewer tickets, fewer humans doing the mundane work. Even today, if you reverse engineer thousands of trouble tickets, you see that for these guys it’s Groundhog Day; they solve the same problems over and over again. Over the last 30 or 40 years, the cost and complexity of managing Windows endpoints specifically hasn’t got that much better. We have more tools that are collecting more data, creating more noise – not always signals – and the ratios of people that are managing these end points hasn’t improved that much. This is the quandary for many of these enormous endpoint management companies and this is why we’re staking our claim that DEX is the keyhole into autonomous endpoint management. That is the category we want to lead,” he said. live remote management, which enables someone on a support desk to log on to a device in real time to remediate and resolve issues, will pit ControlUp against established endpoint management companies. “People may not be cognizant of how disruptive this could be to some of the legacy endpoint management tools and techniques. In the next 12 to 24 months, this is going to take on a life of its own as people see those durable, repeatable use cases and start to recognise the ROI. It’s going to be profound.” Existing endpoint management leaders are also enhancing automation with agentic AI, but in Ayres’ opinion Controlup has the edge thanks to its speed advantage and rich data set. “You see legacy endpoint management tools quickly trying to add AI, but they don’t have as rich a data set as we do. Because of VDI being so wobbly, we set up this three-second clock speed, making us 20 times faster than other VDI players. Our stack of data goes quite a bit higher, wider and faster than others.” Those others include a wide spectrum of companies, from vibecoded start-ups to billion-dollar legacy endpoint management companies. ControlUp is positioned between the two and hoping to bring the best of both to the table thanks to a sizeable base of enterprise customers that have generated a lot of data for its AI model to work with and an agile development culture. “The question for the large legacy endpoint management companies that have way more customers than we do is can they pivot quickly enough to this new era? I’ll put my money on ControlUp right now. The speed at which we’re responding to requests from customers, the speed at which we’re innovating is unlike any company I’ve ever sat in. About 70% of our code today is being written by AI. We’ve just brought a product to market called DaaS IQ that will compete against Hydra and Nerdio in the AVD (Azure Virtual Desktop) optimisation space, and we did that with a very small team in exactly six months. That’s going to be more common. We should be at 90% of our code being written by AI by the end of the year.” chains of this model entirely with one that is able to observe, to act autonomously, to verify and to learn and build on itself,” he said. “I think every manual fix today is going to be something that we can automate out. This is going to free up IT from doing the same repetitive fixes over and over again. In three years, I would suggest, doing it the way we’re doing it today will be a way to lose your job. The goal is not better troubleshooting, it’s no troubleshooting.” In this respect, AEM turbocharges the benefits that ControlUp already offers through what it calls ‘the three Cs’: Collect – gathering data every three seconds on up to 10,000 metrics; Connect – joining the dots by bringing that data together in dashboards and integrating it into other systems; and Correct – remediating issues and vulnerabilities. Collectively, these deliver real-time visibility into VDI, SaaS applications and desktop environments and the automatic remediation of problems, resulting in reduced downtime, fewer tools, fewer agents, fewer support desk tickets and less escalation, all of which carry a cost, while also enabling more informed decision-making relating to software asset management, software reclamation and device life cycle management. “We have a lot of data. We have 10,000 metrics that we’re collecting every three seconds, and in many cases, we’re holding that data set for a whole year. Where we’re seeing most value is when we intersect that data with new AI models that enable us to see patterns that cause anomalies and trouble tickets and productivity loss for the end user and IT teams. Using these models, one, to identify the pattern and the anomaly and, two, to identify very quickly how to automate that so that when it does appear it’s eradicated using a known script or automation is where we’re expending all our energy right now,” explained Ayres. Disrupting endpoint management He adds that this capability and the launch of new features like
RkJQdWJsaXNoZXIy NDUxNDM=