Yorkshire Building Society Group HQ

How Yorkshire Building Society is using AI to give colleagues more time for members

At a internal Data & AI festival in Bradford, Yorkshire Building Society brought colleagues together to explore how AI and agents could reduce admin, build confidence and improve customer service – while keeping people firmly in control.

At Yorkshire Building Society’s head office in Bradford, some of the newest “colleagues” answering customer queries go by the names Penelope, Sam and Alf.

They aren’t people. They’re AI agents.

With more than three million members and customers, Yorkshire Building Society handles a huge volume of enquiries every day, many of them difficult to answer simply.

“A typical day without the use of AI within customer relations can be very, very complex,” says Polly Conner, senior manager of customer relations at YBS. “There’s substantial regulation required within complaints handling, which means there are a lot of admin tasks to do.”

“This technology is going to give our members more choice and better access to information when they need it most“
Simone Fox, Director of Customer Support, YBS

That can mean switching between multiple systems, searching for policies, summarising long histories, and drafting detailed responses before colleagues can focus on the conversation itself.

As a mutual organisation, YBS also knows trust matters. Any use of AI has to be responsible, transparent and supportive, with human judgement always the governing principle.

“This technology is going to give our members more choice and better access to information when they need it most,” says Simone Fox, Director of Customer Support at YBS. “But there will always be a human available to speak to them when they need it.”

People‑first approach

That mindset has shaped YBS’s approach to data and AI over several years. Before rolling out new tools, the organisation invested in modernising its data and technology estate. This included building a cloud-native data platform with Microsoft Fabric and strengthening data governance through Microsoft Purview. It also enhanced its security with Microsoft Sentinel, progressed its infrastructure foundation in Azure, and introduced Windows 365 remote desktops to enable secure, flexible working.

“Our ambition with data and AI is to improve our organisation, help our colleagues be more efficient, make better decisions, and ultimately, serve our members better,” explains Rebecca Fitzgerald, YBS Director of Data and AI.

“All built on trust, with responsibility built in.”

“We’re able to minimise the amount of time staff are spending on admin tasks and actually spend that time talking to our customers“
Polly Conner, Senior Manager of Customer Relations, YBS

The festival brought colleagues from across the business together to learn, experiment and share practical examples of how AI could support their roles.

“I just learned some quick hints and tips that helped me realise there are actually much simpler ways of getting what I want out of building agents,” says festival attendee Jordan Peel, a learning solution designer at YBS.

Front line changes

YBS is rolling out a new customer service platform built using Microsoft Dynamics 365 Contact-Centre-As-A-Service, which brings customer history, previous self-service activity and relevant guidance into one place for colleagues.

An AI assistant can summarise cases, highlight relevant knowledge and help draft responses – reducing the need to search across multiple systems.

“The platform is really going to help our colleagues better serve our members,” says Fox. “It’s going to enable them to have a 360‑degree view of our members, so they’re not having to move between various platforms.”

“They’ve got information at their fingertips,” Fox adds, “which means they can help members in a far more timely fashion.”

Real impact

The three AI agents on show at the festival support different parts of that journey. Penelope helps draft final responses for complaints, and Alf supports this process by searching policies, procedures and past cases. Sam summarises long or detailed complaints.

So far, Sam is the most‑used agent, saving an estimated seven minutes per use. Penelope, used for more complex responses, can save up to 26 minutes at a time.

Yorkshire Building Society branch
The Yorkshire Building Society Group has more than three million members and customers

“That time really adds up,” says Conner. “We’re able to minimise the amount of time staff are spending on admin tasks and actually spend that time talking to our customers, helping them resolve the challenges they’ve faced.”

For YBS, the goal isn’t speed alone. It’s also clarity, consistency and fairness, especially in situations where members may be under financial pressure.

“Being able to create communications that anybody can understand is really important,” Conner adds.

Beyond customer service

Fraser Ingram, Chief Operating Officer at YBS, says the timing of the festival reflects a broader inflection point for the organisation.

“We’ve got the right foundations, the right motivation, and a good understanding of the risks and opportunities of what Copilot can do for us,” he says. “This is now the time we can really drive benefit from it.”

Potential benefits include faster mortgage decisions, better checks on income and expenditure, and more consistent service across the business.

“Speed and consistency will be two key AI-delivered benefits for our members,” Ingram says.

Elsewhere, YBS is piloting AI agents to support internal risk and control testing. Early results suggest efficiency savings of around 40% – freeing up time for professional judgement rather than replacing it.

“I think of Copilot as an intelligent graduate with broad knowledge,” says Pete Balmforth, senior manager of operational risk controls and assurance at YBS.

“It’s powerful, but it’s not a substitute for judgement.”

Building confidence

YBS has also invested heavily in skills. Its Data and AI Academy, launched in April 2024, has already seen 634 colleagues complete more than 1,500 learning modules. Thirty-three colleagues have completed data apprenticeships, with more than 75% achieving distinction.

“It’s about skills, confidence and learning,” Fitzgerald says. “It always comes back to humans making decisions and using technology to support them.”