How a UK Bank Reached 94% Satisfaction Using an AI Assistant. Virgin Money's Redi Case

How a UK Bank Reached 94% Satisfaction Using an AI Assistant. Virgin Money's Redi Case
115
Apr 07, 2026

Virgin Money's AI assistant, Redi, has achieved a 94% customer satisfaction score from over 2 million interactions in just 15 months. Unlike typical chatbots, Redi integrates with over 50 APIs to perform real banking tasks like adjusting credit limits, ordering replacement cards, and setting up direct debits - all within the chat interface. This deep functionality enables 57% of customer queries to be resolved without human intervention during peak times, allowing agents to focus on complex cases.

Key achievements include:

  • 91% of customers successfully resolved their inquiries.
  • 84% of users find the interface easy to use.
  • Winning the "Best Application of AI in Financial Services" award in 2024.

Built using Microsoft Copilot Studio and IBM Consulting, Redi uses natural language understanding to deliver accurate and reliable customer experiences. It has also expanded from credit cards to personal accounts, with plans for more personalized services using generative AI. Virgin Money’s success offers a clear example of how AI can transform customer service in banking.

Why Virgin Money Needed an AI Assistant

Virgin Money

In 2021, Virgin Money introduced a new credit card, but something unexpected happened: customers continued calling the contact center, even after downloading the bank’s feature-packed mobile app. The app was designed to handle many common banking needs, yet people still turned to phone support for basic questions.

This created a major bottleneck. Contact centers were overwhelmed with repetitive queries about app features, card replacements, and credit limits. Agents spent much of their time answering the same questions over and over, which slowed response times, frustrated customers, and left employees feeling drained. To make matters worse, the bank’s existing chatbot wasn’t up to the task. Its limited natural language processing (NLP) capabilities often misunderstood key terms. In one particularly awkward instance, the chatbot misread the word "Virgin" in the company name as offensive language and reprimanded users, alienating them further.

These challenges highlighted the need for a smarter, more customer-friendly solution.

Customer Pain Points

Customers encountered three primary hurdles when seeking support:

  • Limited in-app guidance: Despite the app’s robust features, users struggled to find straightforward help.
  • Restricted support hours: Assistance was only available during standard business hours, leaving customers stranded during evenings and weekends.
  • A lack of warmth in digital interactions: The app felt impersonal, making it difficult for customers - especially those less tech-savvy or financially confident - to navigate without human interaction. While functional, the app lacked the approachable and empathetic feel associated with the Virgin brand.

"We know how important it is for customers to experience digital journeys that feel human, not sterile and functional", said Allegra Patrizi, Managing Director – Business and Commercial at Virgin Money.

These pain points made it clear that a more intuitive and human-like solution was necessary to meet customer expectations.

What Virgin Money Wanted to Achieve

Virgin Money outlined specific goals for its new AI solution. First and foremost, they wanted to provide round-the-clock support capable of handling real banking tasks - not just answering FAQs. By integrating the AI directly with core banking systems through APIs, customers could complete actions like ordering replacement cards or adjusting credit limits without needing to call.

Another key objective was to reduce the burden on human agents, allowing them to focus on more complex or sensitive customer needs. This approach aimed to improve both customer satisfaction and employee morale. Finally, Virgin Money wanted to create a digital experience that reflected the brand’s friendly and approachable personality. The goal was to make digital support feel as helpful and warm as speaking to a live representative, all while being available 24/7.

The takeaway: These goals were designed to tackle the core challenges head-on, paving the way for an AI solution that could transform customer and agent experiences alike.

How Redi Works: Technology and Development

Redi

Virgin Money designed Redi with power and scalability in mind. The assistant operates on Microsoft Copilot Studio (formerly Power Virtual Agents), paired with Bot Framework Composer and Natural Language Understanding technology. This combination was carefully selected to align with the bank's AI-first strategy.

What makes Redi stand out from standard chatbots is its seamless integration with Virgin Money's core banking systems. Through over 50 APIs, customers can perform real-time banking tasks like updating credit limits, setting up direct debits, or ordering replacement cards - all within the chat interface. To ensure security, the entire system is hosted in Virgin Money's private cloud environment.

Redi was also designed to feel approachable and human. Its Natural Language Understanding enables it to interpret casual language, respond empathetically, and even add a touch of humor when appropriate, reflecting Virgin Money's brand identity. Since its launch in March 2023, Redi has managed more than 2 million interactions and can handle over 100 types of queries. This technical foundation allows for quick and efficient real-time processing.

Redi's Technology and Capabilities

Redi's architecture takes it far beyond simple Q&A functionality. Built on Microsoft's Copilot suite, the assistant leverages Power Automate to connect the chat interface with back-end banking systems. For example, if a customer requests a credit limit increase, Redi doesn't just explain how to do it - it processes the request instantly using its API connections.

Redi also initiates proactive communication. For instance, when a customer's card is about to expire, the assistant contacts them to confirm their address before sending a replacement. This approach not only reduces fraud risk but also automates routine tasks, leading to a 97% journey completion rate for these interactions.

Training Redi to understand natural language was a priority. It doesn't rely on specific keywords but instead recognizes everyday language and straightforward requests. This focus on user-friendly design has resulted in an 84% ease-of-use rating from surveyed customers.

"We chose Copilot Studio primarily for the powerful capabilities it offers through the Power Automate platform. What started as basic chatbots in our virtual agents has evolved into a fully integrated AI solution", said Ruaridh Wallace, BizApps Centre of Excellence Lead at Virgin Money.

Working with IBM and Microsoft

Virgin Money partnered with IBM Consulting to shape its AI strategy and build Redi. IBM brought industry expertise to the table, while Microsoft provided technical insights. The project was rolled out in phases, starting with credit card users in March 2023 and expanding to Personal Current Account holders by June 2024.

This collaboration was key to avoiding common issues. IBM's team helped Virgin Money tackle the challenges of integrating AI with legacy banking systems while maintaining regulatory compliance. The effort was led by Adam Paice, Head of Digital Proposition at Virgin Money. This teamwork played a major role in achieving high customer satisfaction rates.

"We have benefited from a truly exceptional IBM support team on the ground to steer us away from common mistakes that are made when deploying these services", Paice noted.

The results speak for themselves. When Redi launched for credit card customers in March 2023, it immediately handled 51% of all credit card-related conversations, with over half resolved without human assistance. By June 2024, the same Microsoft-driven framework was extended to current account customers, providing consistent 24/7 support across multiple banking services.

Virgin Money's decision to combine Microsoft's technology with IBM's expertise laid the groundwork for a platform that evolved from basic chatbot functionality into a full-fledged conversational AI solution for banking.

The Results: What Redi Achieved

Virgin Money Redi AI Assistant Performance Metrics and Key Results

Virgin Money Redi AI Assistant Performance Metrics and Key Results

With its well-thought-out design, Redi has proven itself as a game-changer in customer service. Virgin Money's decision to integrate Redi has led to notable improvements in both customer satisfaction and operational performance.

Customer Satisfaction Numbers

Redi's impact is clear through its 94% customer satisfaction score and 57% containment rate during peak times, with 91% of customers successfully resolving their inquiries. Virgin Money internally refers to these metrics as "smile scores", and Redi has achieved one of the highest ratings within the bank's operations.

The numbers speak for themselves: 84% of users found Redi's interface easy to navigate. These results highlight how Redi's seamless API integrations and proactive features contribute to outstanding customer experiences.

Performance Metric Result
Customer Satisfaction Score 94%
Goal Achievement Rate 91%
Peak Containment Rate 57%
Ease of Use Rating 84%
Total Interactions (Mar 2023–2024) 2+ million

How Redi Helps Human Agents

Rather than replacing human staff, Redi works alongside them, enhancing their ability to handle complex and sensitive cases that require empathy and expertise. When a situation calls for a human touch, Redi ensures a smooth handoff by transferring all necessary context to the agent.

"My team sees Redi as another colleague. It's the one who passes a customer to them when a human touch is needed."

  • Alison Lattimer, Head of Customer Service Operations, Virgin Money

Redi also improves efficiency through proactive messaging. For instance, when customers receive reminders like card expiry alerts, 54% engage with the notifications, and an impressive 97% of those who engage complete the entire digital process. This approach reduces the volume of routine inbound calls while maintaining high service completion rates.

These enhancements not only support the customer service team but have also earned industry recognition.

Industry Recognition

Virgin Money's collaboration with Redi has been celebrated with the "Best Application of AI or ML in Financial Services" award at the 2024 Card & Payments Awards. This recognition underscores the bank's success in deploying conversational AI within a regulated industry.

"Our customers tell us how much they enjoy interacting with Redi in the credit card app, and to receive industry recognition of our hard work is the cherry on top."

  • Adam Paice, Head of Digital Proposition, Virgin Money

The award highlights both the cutting-edge implementation of this technology and its tangible benefits for customers. Redi has shown that AI assistants can meet customer needs while delivering satisfaction levels that rival human interactions.

Key Takeaway: Virgin Money's standout results - 94% satisfaction, 57% containment, and 91% goal achievement - demonstrate how thoughtfully designed AI tools can create meaningful support for both customers and staff.

What Other Banks Can Learn from Virgin Money

Virgin Money's approach offers a roadmap for banks aiming to implement AI effectively. Their success with Redi highlights how thoughtful design and seamless integration can lead to high customer satisfaction. By focusing on customer-first solutions, Virgin Money demonstrates how AI can truly enhance the banking experience.

Connecting AI to Core Banking Systems

One of the standout lessons from Virgin Money is the importance of deeply integrating AI with core banking systems. Redi, their AI assistant, connects to these systems through over 50 APIs, enabling tasks like credit limit adjustments and card orders. Laura MacLeod, Product Strategy Senior Manager at Virgin Money, described the transformation:

"We transitioned from outdated chatbots to a unified conversational banking experience."

Virgin Money adopted a phased rollout strategy, starting with credit cards and later expanding to current accounts. This allowed them to fine-tune the technology before scaling it across their services. The result? A reusable platform capable of adapting and growing with the bank's needs.

The benefits of this integration are clear. For instance, 97% of customers who engage with Redi's proactive messages complete their entire digital journey. This shows that when AI provides end-to-end solutions, customers are more likely to engage and complete their tasks. Beyond efficiency, this level of integration lays the groundwork for building stronger customer relationships.

Building Customer Trust in AI

Earning customer trust in AI requires transparency and a human touch. Virgin Money prioritized creating a relatable and empathetic tone for Redi, ensuring interactions felt personal and approachable. Allegra Patrizi, Managing Director of Business and Commercial at Virgin Money, underscored this focus:

"We know how important it is for customers to experience digital journeys that feel human, not sterile and functional".

Transparency was a key element in their strategy. Redi is designed to complement human agents, not replace them. When a situation calls for empathy or specialized knowledge, Redi hands off the customer to a human agent, complete with all the context needed to ensure a smooth transition. This approach reassures customers that they won't be left navigating complex issues alone.

Proactive security measures further strengthen trust. For example, when a customer's card is about to expire, Redi sends a message to confirm their delivery address before sending a replacement. This feature, which achieved 54% customer engagement, not only reduced fraud risks but also demonstrated the bank's commitment to adding real value to its customers' lives.

These strategies offer a clear blueprint for other banks looking to enhance their AI systems. By focusing on integration, phased rollouts, and human-centric design, banks can create AI solutions that are both effective and trustworthy.

Key Takeaway: To successfully implement AI in customer service, banks should focus on deep integration with core systems, refine their platforms through phased rollouts, and ensure AI interactions are transparent and empathetic, with seamless handoff protocols to human agents. This combination builds both efficiency and trust.

What's Next for Redi

Virgin Money is taking Redi to the next level by incorporating OpenAI and Microsoft generative AI to create more personalized and context-aware customer interactions. The goal is to make these conversations feel natural and tailored, enhancing the overall banking experience.

A major upgrade in the works is combining credit card and current account functionalities into a single virtual assistant. This unified approach eliminates the need for separate interactions across different banking products. Instead, customers will engage with one AI assistant that understands their full banking relationship.

Adam Paice, Head of Digital Proposition at Virgin Money, emphasized their thoughtful strategy:

"Our partnership with IBM has helped us to get the most out of the suite of Microsoft Co-pilot products and find the balance between innovation and control".

Striking this balance is critical. While generative AI can make banking interactions more conversational and helpful, ensuring the accuracy and security of responses remains a top priority. Virgin Money's phased rollout - starting with credit cards, then moving to current accounts, and now integrating generative AI - allows them to fine-tune the technology before expanding its scope.

This evolution aligns with Virgin Money's upcoming business transfer to Nationwide Building Society, set for April 2, 2026. This transition opens the door for further advancements in digital banking. Their commitment to becoming the leading digital bank in the UK continues to drive these innovations, prioritizing both customer trust and forward-thinking technology.

Key Takeaway: Virgin Money is leveraging generative AI to deliver more natural and personalized interactions while unifying multiple banking services under one assistant. Their partnership with IBM ensures they maintain security and control as they advance conversational banking.

Conclusion

Virgin Money's journey with AI demonstrates how thoughtful integration can reshape customer support in banking. Their AI chatbot, Redi, achieved an impressive 94% customer satisfaction across more than 2 million interactions, handling over 50% of credit card inquiries independently [2, 3]. These results highlight how customers respond positively to AI solutions that feel approachable and work seamlessly with existing banking systems.

What Made Redi Stand Out?

One major factor was deep API integration, which allowed Redi to go beyond answering basic FAQs. Tasks like adjusting credit limits and ordering new cards became possible, transforming Redi into a practical digital assistant, not just a chatbot.

This approach didn't just benefit customers - it also supported human agents. By taking care of routine questions, Redi gave staff more time to focus on complex, high-stakes issues that require a personal touch.

"My team sees Redi as another colleague. It's the one who passes a customer to them when a human touch is needed."

  • Alison Lattimer, Head of Customer Service Operations, Virgin Money

This hybrid model improved efficiency while enhancing job satisfaction, proving that AI and human teams can work together harmoniously. The industry has taken notice, too - Redi earned the "Best Application of AI in Financial Services" award at the 2024 Card & Payments Awards [2, 3], setting a high bar for conversational banking.

Lessons for Other Financial Institutions

Virgin Money's phased approach offers a clear roadmap for banks looking to implement AI chatbots. Starting with a single product line, focusing on strong system integration, and developing an approachable brand voice were key steps in their success.

For organizations exploring AI solutions, the takeaway is clear: prioritize integration over surface-level functionality, design for empathy as well as efficiency, and see AI as a partner for your human team - not a replacement. Virgin Money's success story provides a blueprint for driving innovation in digital banking support.