How Burns & McDonnell Unites Nexthink’s Full Stack to Transform The IT Support Experience
For most IT organisations, the service desk begins when an employee reports a problem. At Burns & McDonnell, the goal is increasingly to resolve the issue before that conversation is ever needed.
By the time an engineer or project team member asks for help, valuable time has often already been lost. Supporting engineers, architects, and construction professionals around the world, the team had already embraced DEX to understand what was happening across devices, applications, and employee experiences. The next step was to move beyond faster response and reduce the number of issues employees experienced in the first place.
Nexthink Spark made that ambition more practical by combining real-time endpoint telemetry, enterprise knowledge, AI-driven diagnostics, and governed remediation into a support experience that could resolve more issues at the point of need. Employees could describe a problem in natural language, Spark could investigate what was happening on the device, and approved actions could be taken without waiting for a service desk analyst.
The goal was not to replace IT, but to create a support model where routine issues are handled automatically wherever possible, employees stay focused on project work, and IT gains more time to improve the digital workplace instead of repeatedly fixing the same problems.
“I really see us being able to resolve upwards of 80% of the issues that occur within our IT environment before a customer has to report those.”
Looking beyond the ticket
That ambition was shaped by the way employees experienced technology problems long before IT became involved. By the time someone submitted a ticket, they had often already spent valuable time trying to solve the issue themselves, interrupting work and delaying projects before eventually deciding they needed help.
For Burns & McDonnell, the cost of that delay is measured less by the ticket itself and more by the productive engineering time that disappears while employees wait for a resolution. Engineers routinely work with highly specialised applications, enormous datasets, and complex design environments where workstation performance directly affects project delivery. Matt Harmon, Director of IT at Burns & McDonnell describes scenarios where point cloud data measured in terabytes could take hours to process, only for an unexpected issue to force the entire task to begin again. Losing several hours in the middle of a project is not simply frustrating. It can push deadlines, interrupt client work, and reduce valuable engineering time.
That understanding changed how the team viewed success. Rather than treating tickets as the primary measure of IT performance, they began focusing on the disruption employees experienced before a ticket was ever created. If those interruptions could be identified and resolved earlier, employees could continue working without ever entering the traditional support process.
Building on a proactive foundation
Long before introducing Spark, they had invested in proactive monitoring and automated remediation through Nexthink. The organisation was already identifying common endpoint issues, triggering automated actions, and preventing employees from experiencing problems that would previously have generated support calls.
The results demonstrated how much value prevention could create. In a single 30-day period, the team estimated that those proactive capabilities saved approximately 5,000 hours by detecting and resolving issues before employees reported them. Instead of waiting for someone to call the service desk, many problems simply disappeared before they became visible to the user.
In areas where automation has been introduced, Burns & McDonnell has cut year-over-year ticket growth from around 30% to just 2%. That means the team can handle more support demand without ticket volumes rising at the same pace.
This progress ultimately changed how the team thought about support. With automation already preventing a large proportion of repetitive incidents before employees even noticed them, the team could turn its attention to helping employees resolve the issues that still required interaction with IT without unnecessary delay.
Moving from scripted automation to intelligent resolution
The organisation had already introduced employees to virtual IT support through Dottie, (their in-house agent) but like many first-generation chatbot experiences, it still relied on predefined keywords, scripted paths, and carefully maintained workflows. The automation worked, but employees often had to describe the issue in the right way before the correct action could be triggered.
With Spark behind the experience, support no longer had to depend on employees navigating menus or matching predefined intents. Spark could interpret requests in natural language, combine the conversation with live endpoint telemetry, organisational knowledge, and approved remediation actions, and investigate what was happening on the employee’s device before deciding how to respond.
Because employees already knew Dotty, the team did not need to introduce a new support experience. It evolved the platform into Dottie Fix-It, with Spark providing the intelligence behind the scenes. Employees continued using a tool they already trusted, while Spark added deeper diagnostics, recommended next steps, and, where appropriate, executed approved remediation automatically.
That allowed the team to build on existing adoption instead of replacing it, turning a familiar virtual agent into a more capable AI support experience.
Winning trust through real outcomes
As with any AI deployment, the team expected a degree of scepticism from employees. Many questioned whether an AI assistant could genuinely diagnose technical issues without human involvement, particularly in an engineering environment where specialist applications are central to day-to-day work.
On the first day Spark was made widely available, it resolved 120 issues without escalation, exceeding the number of issues handled by the existing service desk that day. Current Spark data shows 69% of interactions were recorded as resolved, 13% escalated and 18% abandoned.
Confidence changed quickly once employees experienced successful resolutions for themselves. One engineer struggling to open a project file asked Dottie Fix-It for help. Spark investigated the issue, executed the appropriate action, and restored access immediately. The service desk could have solved the issue, but Spark changed the experience for the employee. There was no need to wait in a queue, explain the problem to another person or hope someone was available.
Another employee preparing for a presentation while travelling overseas was able to resolve a PowerPoint issue in the middle of the night without waiting for the service desk to come online. Instead of delaying important client work until business hours, Spark provided immediate assistance regardless of location or time zone.
For employees, this can make the difference between a technology issue becoming a blocker and being resolved quickly enough for work to continue. At the same time, routine problems are kept away from the service desk, supporting the wider model Burns & McDonnell is building.
Helping IT spend more time improving IT
As more routine support has been automated, Burns & McDonnell has already moved people off the service desk and into roles focused on DEX and automation. They can now spend more time building new automations, improving proactive support and reducing the amount of repetitive work reaching the service desk.
Matt describes this evolution as moving from "human in the loop" to "human on the loop." Instead of participating directly in every support interaction, analysts increasingly supervise automation, review outcomes, refine workflows, and identify new opportunities for improvement.
A vision measured by prevention
Burns & McDonnell’s longer-term goal is for 80% of issues to be resolved proactively before employees need to report them, another 15% to be handled by Spark, and just 5% to require direct human support. This would take much of the repetitive Level 1 work out of the service desk, leaving the team to handle the issues that need their expertise.
As more issues are prevented or resolved without the service desk, the team is also looking beyond traditional measures like response times and ticket volumes. They are paying more attention to the time returned to employees, the disruption avoided and the impact on the wider business.
Key takeaway
Burns & McDonnell’s progress shows how targeting the routine issues creating the most support demand can start to change the role of the service desk. Identifying which of those issues can be prevented, diagnosed or resolved automatically gives employees a faster route back to work while freeing IT to focus on problems that need human expertise.
Spark helps put that model into practice by bringing real-time context, diagnosis and remediation into the support experience, so more issues can be resolved before they need to reach the service desk.