CDW Builds an AI Chatbot To Automate Help Desk Tickets for Its Own Employees
As remote and hybrid work reshaped coworker expectations, CDW used Moveworks-powered artificial intelligence to simplify access to support, expand self-service and create capacity for higher-value technical work.
For enterprise IT teams, the pressure is familiar: more tickets, more complexity and not enough hours in the day. The answer isn’t always more head count. Sometimes it’s smarter technology, deployed in the right place.
Like many organizations during the pandemic, CDW was forced to adopt a remote hybrid model when it was largely an in-office company. The move upended the help desk dynamic and forced the support team into a new normal with a heavy pace.
The company had an accessible knowledge base in ServiceNow, but employees weren’t using it to its full potential. Remote work reshaped how workers expected to get help, accelerating their digital dependency and raising expectations for instant answers.
That shift hit the tech support team directly. Pre-pandemic, the help desk metrics focused on talk time and answer percentage. “If they hit that talk time limit, they were able to send somebody deskside to support the employee,” says Andrew Mikulec, a senior manager on the support and operations technology service desk at CDW.
After the pandemic, deskside support disappeared as an option. The key performance indicator shifted to first contact resolution. But that’s hard to do with a small service desk team.
At the same time, IT issue volume surged.
“Talk time basically doubled when they went remote, and the answer percentage probably decreased by 25%,” notes Mikulec. Problems were more complicated, and the usual recourse of having someone physically stop by and walk the employee through the solution was no longer available.
Compounding the problem was a generational shift. The company’s ranks increasingly included workers who had grown up in the digital age.
“Coworker expectations changed,” says Osayande Myhand, director of technology support services at CDW. “These workers expect to have answers at their fingertips. People increasingly expected workplace technology to be as intuitive and responsive as the digital services they use every day. We recognized that our support model needed to evolve accordingly, giving coworkers immediate access to help while allowing our technical teams to focus on issues that require deeper expertise.”
The knowledge base alone could no longer meet the need. “You had to go find it, you had to search through it, click through an article, read halfway through an article before you figured out that it wasn’t the article you were looking for,” says Mikulec.
Leaders faced a clear structural constraint: The IT team had no interest in adding yet another portal to the mix. Workers already toggled between too many systems throughout the workday, losing momentum with every switch. The business could see the productivity costs.
The solution had to live inside the tools employees already used every day. With that in mind, CDW implemented an AI chatbot inside Microsoft Teams that is powered by Moveworks, which was acquired by ServiceNow.
“The chatbot made sense because it was an easy-to-use tool that we could stick right where people were already working, and it could help tackle a lot of those simple, repetitive questions,” explains Mikulec.
Embedding an AI Chatbot Where Employees Already Work
The goal was a “Level 0” fix: a self-help solution that required no human intervention.
“We needed to provide coworkers with an intuitive self-service experience that could deliver immediate value while reducing unnecessary friction in the support process,” says Mikulec.
While a knowledge base assumes employees know where to look, “an AI chatbot allows the coworkers to simply ask a question,” says Myhand. “They don’t have to know where to look, because it’s just a one-stop shop, an intelligent digital front door that can resolve issues before they ever become tickets.”
“We had already invested in mature knowledge management and established service management practices,” Myhand says. “That foundation allowed us to focus on expanding how coworkers accessed support, rather than rebuilding the underlying processes.”
To bring that solution to life, “we looked at several vendors, and after doing some bake-offs, we chose Moveworks, now ServiceNow,” Mikulec says. The company offers an AI Assistant platform that empowers workers to search and act across business applications.
Drawing on its practitioner perspective to deploy this technology at scale, CDW guided the first phase in identifying which systems the chatbot, named Harold, would connect to. Starting with ServiceNow gave employees immediate access to the knowledge base. From there, the team extended the chatbot’s capabilities to take action, not just surface information.
“We connected it to our Active Directory environment so that if someone’s account got locked while they were working, or they needed to reset their password, they could use the chatbot for that,” Mikulec says. “They’d be able to get that done automatically, without needing to pick up the phone and call the service desk or open a self-service ticket.”
“The goal wasn’t to replace the service desk. The goal was to elevate it.”
— Osayande Myhand, Director of Technology Support Services, CDW
Keeping adoption frictionless meant a critical design choice. “One of our guiding principles was to meet coworkers where they already work,” Myhand says. “Embedding Harold in Microsoft Teams reduced friction and made support part of the normal flow of work rather than another destination coworkers had to learn.”
With IT support embedded in Teams, workers didn’t have to change a thing about how they worked. “It was literally right there, in the tool that they used all the time,” he says.
To ensure successful adoption, the team put a heavy emphasis on change management. “We approached adoption deliberately,” Myhand adds. “We were transparent about the capabilities available in the initial phase, demonstrating where Harold could create immediate value and build confidence through practical use cases. That allowed us to expand responsibly as adoption, content and integrations matured.”
At first, it was strictly an IT tool. “We were using IT for knowledge. It would search those knowledge bases within ServiceNow. It could open a ticket, give you a status update,” Mikulec says. “It could unlock your network account if it was locked out, and it could reset your password if you needed to do that.”
As adoption grew, so did the chatbot’s capabilities.
Mikulec described how the platform evolved beyond IT support. “It gradually grew over time. We added a communications module, so that we could say, ‘Hey, at this time, on this date, I want you to send this message to everyone.’ It broadened the communication strategy around things like maintenance and outages. After years of sending emails about things like that, it offered a fresh, new way to give people the information they need without filling up their inboxes,” he says.
The Benefits of AI-Powered IT Support
$1.84 vs. $13.50 The median cost of an IT issue resolved via automated self-service or conversational AI versus via a live human agent-assisted ticket, roughly seven times cheaper1
43% The percentage of IT executives who point to IT operations as a primary source of daily digital friction and lost worker productivity2
80% The percentage of common customer service issues that will be resolved by agentic AI autonomously by 2029 without human intervention, leading to a 30% reduction in operational costs3
10.5 hours The number of hours per week, per person, by 2030 that help desk support specialists stand to save as AI assumes more routine operations, with nearly four-fifths of that shift driven by agentic AI alone4
Sources: 1digitalapplied.com, “One Customer, Two Logins, Three Domains: Portal Sprawl,” Aug. 12, 2026; 2CDW, 2026 CDW Frictionless Enterprise Research Report, April 2026; 3gartner.com, “Gartner Predicts Agentic AI Will Autonomously Resolve 80% of Common Customer Service Issues Without Human Intervention by 2029,” March 5, 2025; 4ServiceNow, “AI will transform the role of the help desk support specialist,” Sept. 16, 2025
Thousands of Chatbot Interactions, Fewer Service Desk Calls
The results came quickly. It took IT roughly 20 hours to launch the cloud-hosted solution, a significant win for time-to-value, as the ROI was immediate. In the first 90 days, the business saw 331,000 interactions among 18,000 users and had a 58% engagement rate. Of those, 90% were successful interactions, and the chatbot earned a 4.3 out of 5 customer satisfaction score, a strong signal that employees weren’t just using it, they were pleased with the results.
With the chatbot in place, even remote workers get instant support. The system no longer requires submitting trouble tickets, waiting for a callback from the service desk, and then waiting again for a technician. The entire process is streamlined.
“With that intelligent digital front door, coworkers can often resolve issues immediately, without navigating multiple support layers or waiting for assistance,” says Myhand. “The result is a more seamless support experience and greater capacity for our teams to focus on higher-impact activities.”
The chatbot now accounts for 16% of all contacts to the service desk, on average. And 80% of employees now turn to the chatbot when they need help. The number of service calls dropped substantially. “The chatbot creates capacity equivalent to a meaningful portion of the service desk workload, allowing technicians to focus on requests that benefit most from human expertise,” says Mikulec.
For the IT team, that freed up capacity for higher-value work.
“The goal wasn’t to replace the service desk. The goal was to elevate it,” says Myhand. “Intelligent self-service allows our teams to focus more of their expertise on complex issues, meaningful coworker interactions and improvements that strengthen the broader technology experience.”
To sustain that momentum, CDW upskilled a portion of its IT technicians with AI capabilities, enabling them to manage and improve the chatbot directly. The result was faster updates and a better overall user experience.
With employees solving their own problems, “the others are able to spend more time on phone calls or tickets,” says Mikulec. “For someone sitting a thousand miles away, they’re able to spend the time to get them fixed on that first contact, rather than taking it, escalating it and having to wait to get it assigned to someone.”
Across the broader IT organization, engineers and developers who once spent their time on bug fixes now drive larger projects and initiatives forward.
“Now, they can carry forward on these bigger infrastructure projects,” says Mikulec.
Getting End-User Buy-In Was Key to the Success of the Chatbot
Change management was central to the rollout. Internal marketing campaigns introduced the chatbot to employees, supported by a short video walkthrough explaining how to use it. When employees called in or opened a ticket, the IT team seized the moment.
“We’d take the time to actually show them how to do it through the chatbot, to build that confidence,” Mikulec says.
The biggest lesson? CDW moved cautiously at first, but hesitation costs more than it saves.
“Looking back, this probably could have gone faster,” says Mikulec. “An effort like this is always going to continue to evolve and grow. You can enable it to do minimal things at first, and everything beyond that is enhancement. It becomes a kind of natural marketing campaign: ‘Now we’ve done this, now we’ve done that.’ Then, you can just build on that adoption.”
“Harold is not a one-time implementation. It is an evolving enterprise capability,” says Myhand. “Our focus is to continue expanding the value it provides, improve the quality of each interaction and thoughtfully identify where AI can remove friction across the coworker experience.”
Connect with a CDW expert now to explore how you can accelerate outcomes and improve ROI using Moveworks.
Adam Stone
CDW Expert
