Measuring Workplace AI by Employee Outcomes With DXC Technology

What if the most useful measure of workplace technology was not the number of tickets closed, but the amount of productive time returned to employees?

In this episode of Tech Talks Daily, I speak with Kelly Candler, Global Offering Lead for Workplace and Business Process Services at DXC Technology, about moving digital workplace services away from activity metrics and toward employee and business outcomes. Kelly has worked on both sides of the managed services relationship. Before joining DXC, she was a customer of the company and several of its competitors, giving her a practical view of what large enterprises expect from workplace technology and where support models continue to disappoint.

Kelly begins with a simple distinction. Traditional IT reporting often focuses on ticket volumes, response times, and calls answered. Employees care about whether they can do their jobs, how much time they lose, and how quickly they can return to productive work. When the outcome becomes the starting point, support is judged by its impact on the employee rather than the amount of activity generated behind the scenes.

That matters because many companies already have crowded workplace technology estates. Years of investment have produced overlapping tools, legacy applications, automation, device platforms, and service channels. Adding another AI product can increase cost and confusion if it replaces nothing and connects to little. Kelly argues that organizations should understand the technology they already own, identify the employee outcomes they want, and use orchestration to connect those investments rather than discarding them automatically.

DXC positions its Workplace Services offering as a people-centered, AI-native service enabled by DXC OASIS, its orchestration platform for Human+ agentic AI workflows. In the interview, Kelly describes Human+ as a partnership in which AI handles repetitive work while people retain creativity, judgment, and accountability. The intention is to make support more proactive, provide employees with help through familiar channels, and allow experienced teams to concentrate on work that requires human knowledge.

Device performance provides a practical example. In a reactive model, employees discover that a laptop has become slow, unstable, or unusable after productivity has already been lost. Kelly explains that operational data and automation can identify patterns such as declining battery health, storage problems, and application conflicts before the employee experiences a failure. The system may resolve the issue automatically or arrange a replacement before work is interrupted.

Kelly says DXC performs over 82 million proactive checks and remediations annually across over one million managed endpoints. These are company-reported operating figures, but they illustrate the scale at which preventative workplace support is being applied. She expects agentic AI to extend that model by learning from operational data and improving the speed and scope of proactive action.

We also examine DXC’s reported outcome figures. The company says Workplace Services can produce a 40 percent reduction in operational complexity, 60 percent fewer service desk calls, resolve 50 percent of device issues before employees notice, and return over 15 hours of productivity to each employee every month. Kelly explains that the productivity calculation considers incident resolution, device availability, PC performance, and onboarding. The actual result will depend on the employee persona, environment, baseline, and method of measurement, so leaders should examine how each figure was calculated before applying it to their own workforce.

The discussion then moves from technology to adoption. Kelly says organizations often know where to run an AI proof of concept but struggle to turn the result into a production capability. DXC’s response is an Exponential Blueprint that assesses the environment, maturity, culture, and potential value areas, then helps customers move from a smaller test toward wider use. The wider lesson is that a successful pilot needs a defined route into existing operations, ownership, governance, and measurement.

Governance becomes especially important when Human+ workflows begin making decisions at scale. Kelly notes that human service desk employees already make mistakes, so the standard for AI should focus on risk tolerance, oversight, and what happens when an error occurs rather than assuming perfection. DXC has introduced an AI code of conduct alongside its employee code of conduct. Kelly says humans remain accountable for the actions taken by AI agents, much like the accountable role in a RACI model.

Trust also depends on how leaders explain the changes to employees. Automation can raise concerns about surveillance and job loss. Kelly frames the opportunity as removing repetitive work so people can concentrate on decisions, creativity, and business outcomes. That promise will only be credible if employees understand how the technology is used, what remains under human control, how performance is measured, and where they can challenge an automated decision.

For CEOs deciding where to begin, Kelly recommends support and device management because both affect nearly every employee and can produce measurable results. She also suggests moving toward experience level agreements that connect workplace performance with employee outcomes. The longer-term value comes from understanding the broader environment in which employees work and improving how its parts operate together.

If workplace AI is intended to give people time back, should organizations retire ticket volume as their headline measure and report the employee impact instead? Listen to the episode and share your thoughts.