CASE STUDY: EE

EE is the UK’s largest mobile network operator, serving more than 25 million subscribers across consumer and business segments. When the planning and scheduling team saw that their FlexiTime benefit — already made possible by a consolidated NiCE Workforce Management platform — was still consuming 12 full-time roles and generating consistent employee frustration, they chose to build the solution themselves. The result is a fully automated, 24/7 self-service FlexiApp that eliminated the manual process entirely, grew daily FlexiTime adoption by 580%, and reduced absence in the first live area by 28% year over year.

01 Before

A Manual Process That Could Not Keep Up with the Benefit It Served

EE is the UK’s largest mobile network operator, serving more than 25 million subscribers across consumer and business segments under the EE, BT, and Plusnet brands. Behind that customer-facing scale sits a planning and scheduling operation responsible for keeping service levels intact and colleagues supported across 17,000 customer-facing roles. That operation runs on NiCE Workforce Management, a platform EE consolidated across all three brands in 2022, specifically to unlock greater workforce flexibility. At the heart of that flexibility was a benefit EE had offered for years: FlexiTime.

The proposition was sound. Colleagues could earn time back, trade hours, and manage their working week with greater autonomy. The problem was not the idea. The problem was how it worked.

The process required 12 members of staff rotating through an administrative role, processing requests manually during office hours only. On a busy day, the team handled approximately 650 requests. A colleague wanting to flex on a Friday afternoon waited until Monday. Weekend requests sat untouched.

The system ran on spreadsheets, email inboxes, and human judgment, and it showed. The manual administration layer sat entirely outside NiCE Workforce Management, layered on top of a platform that was never designed to process FlexiTime requests by hand. Wayne Mitchell, Planning and Scheduling Manager at EE and the architect of the team’s automation philosophy, put it directly, “the proposition was great, the implementation was not. It was clunky, full of friction, and the 12 people processing requests were doing nothing but pure transactional work with no value-add to show for it.”

Employee satisfaction surveys consistently flagged FlexiTime as a source of frustration. Colleagues who knew the benefit existed often did not bother using it because the turnaround was too slow and the process too opaque. EE had already invested in workforce flexibility. The employee experience behind it was breaking down. The team could see exactly what needed to change. What they needed was a way to act on it.

02 Desire to change

Move the Skill to Where the Knowledge Lives

The catalyst came from an unrelated decision. A third-party software contract was approaching end of life, and the tool integrated poorly with NiCE Workforce Management. Rather than renew, Wayne Mitchell opened a different conversation: what if EE used the technology it already had to build something better?

EE had been running on the consolidated WFM platform since 2022. Mitchell knew that NiCE’s automation capabilities had been designed as guided help tools for contact center agents and asked whether they could be repurposed for planners and schedulers instead. No one had tried it, but there was no technical barrier. That first project succeeded: a manual scheduling task was replaced entirely by automation Claire Gallagher and Gareth Wilson built themselves. The result gave Mitchell the proof of concept and the organizational credibility to take on FlexiTime. As he put it: “We’ve got this massively friction-heavy process.

We’ve just used NiCE here to really slicken things up and deliver automation. Why can’t we just do it again?”

Behind this was a philosophy Mitchell had been developing for years: move the technical skill to where the functional knowledge lives, rather than the other way around.

Claire Gallagher, a Senior Scheduler, and Gareth Wilson, a Resource Analyst Planner, became the team. Neither had built automation software. What they had was deep knowledge of how FlexiTime worked, where the friction lived, and what a better process would feel like.

03 NiCE solution

Built From Scratch. Shaped by Real People. Eighteen Months in the Making.

The Build: Two People, No Playbook

Gallagher led the robotics and back-end automation, learning from scratch while building in parallel. The experience was, as she described it, like going back to university: “Every single day is a school day. Every day I’m coming in and it’s a school day, but also making progress.” Progress she celebrated in her own way. “I would have what I call desk dances,” she said. “Every time I had a success everybody around me would know about it because I would be on my chair spinning and dancing.” She built something remarkable by refusing to wait until she felt ready.

Wilson focused on the Power Apps front end, translating back-end automation into an interface that 10,000 colleagues could use without thinking about what was happening behind it. As the build evolved, the gap between what they originally planned and what they actually built grew significant. “What we originally thought we were going to build is very different now to what we originally thought it was going to look like,” Wilson said. “It’s been a journey and I’ve learned a lot.” That honest reflection is also what made the end product better than anything a requirements document could have specified.

Rather than automating the existing manual process exactly as it ran, the team used the build phase to improve it. Pilot groups were engaged early, with a dedicated Teams chat running throughout testing so feedback could flow in real time. Issues were identified, investigated, fixed, and republished in cycles measured in hours. That loop surfaced things no design document would have anticipated, including users unfamiliar with 24-hour clock format entering invalid times, which required error handling that only emerged through working with real colleagues in real conditions.

The Architecture: Platform as Foundation

That was possible because NiCE Workforce Management gave them the operational foundation to build on. SmartSync handled schedule write-back. Data Explorer surfaced real-time schedule data to power the app’s rules. Real-Time Designer served as the automation engine connecting every component. EE identified the process problem and designed the change. NiCE provided the platform that made it reliable, scalable, and sustainable.

The technical architecture brought NiCE Workforce Management, Real-Time Designer, Data Explorer, and Microsoft Power Apps together into a single automated workflow. Power Apps provides the interface colleagues interact with: simple, mobile-friendly, completely abstracted from the complexity behind it. NiCE Workforce Management holds and processes schedule changes. Real-Time Designer serves as the automation engine. Data Explorer surfaces real- time availability. These components run together, invisibly, behind a front end that simply works.

Move the technical skill to where the functional knowledge is, rather than trying to move the functional knowledge to where the technical skill is.

Wayne Mitchell

Planning and Scheduling Manager, EE

Change Management: 12 FTEs Returned to Customer-Facing Roles

The 12 staff assigned to FlexiTime administration were drawn from agents on a rotating development opportunity, roughly every three to four months. Automating the process meant those 12 people returned to customer-facing roles. As Mitchell put it: “Those 12 people could go back to saving customers. Not 12 people lose their jobs and exit the company.” That truth changed the nature of the conversation from resistance to buy-in. Being transparent about what automation was for, acting visibly on feedback, and demonstrating that colleagues were part of the build did the rest.

04 Results

A Process That Once Required 12 People Now Requires None

The manual process that required 12 rotating FTEs is fully automated. Daily FlexiTime requests grew from approximately 650 to around 4,000, a 580% increase driven by removing the friction that had suppressed adoption for years. The system now processes requests around the clock, with responses delivered in seconds.

In EE’s largest operational area, the first site to go fully live, absence dropped 28% year over year. Mitchell is careful with that figure: other initiatives contributed and the correlation cannot be fully isolated, but the timing tracks directly with the rollout and the pattern is repeating in subsequent areas. The estimated annual saving from redeployed admin effort runs to approximately £450,000. That was not the goal. Employee experience was the goal. The financial figure followed.

Leadership’s response has shifted from approval to appetite. Requests are coming in for behavioral pattern analysis from FlexiApp data, and some operations leaders are actively considering rationalizing adjacent leave processes now that FlexiTime is efficient enough to replace them. For Gallagher and Wilson personally, new job roles were formally created on the back of the program, and leadership approved a dedicated internal automation team built around the same decentralized model. Across EE, they are known simply as the Flexi people.

05 Future

A Roadmap of Nearly 100 Projects. Version Two Already Underway.

Wayne Mitchell’s team is already building FlexiApp version two, with near-term improvements including near-real-time balance updates and proactive Working Time Directive compliance checks that surface advisory alerts before scheduling issues become problems. Beyond FlexiApp, the team is working through a roadmap of close to 100 automation projects spanning planning, HR, real-time reporting, and training scheduling.

EE made a deliberate decision about where AI fits in that roadmap. For FlexiApp, deterministic automation was the right tool: clear logic, defined rules, no need for AI-powered processing. As the roadmap expands into less structured domains, the team expects that calculus to shift.

The technology is getting better and it’s getting easier to do. If you’ve got a good idea and you believe it will work — just go for it. Version one will not be perfect. That’s not the point. The point is to start the journey.

Wayne Mitchell

Planning and Scheduling Manager, EE