Digital Transformation Consulting

The case study focuses on VOXI, a sub-brand of Vodafone targeting UK consumers aged 16-29, which faced the challenge of engaging Gen Z customers through customer service. Traditional chatbots struggled with understanding complex, multi-query interactions and often delivered responses that felt out of tone—either too juvenile or not professional enough—limiting their effectiveness in representing the brand.

To address these issues, Accenture developed a first-of-its-kind large language model (LLM) AI chatbot tailored specifically for VOXI. Built on the Azure OpenAI framework via Microsoft’s Azure AI Studio, the chatbot was refined over approximately three months through multiple iterations. It was programmed with an intent classifier to filter relevant questions about VOXI mobile services, trained on diverse customer scenarios, including different emotional tones, to enhance response appropriateness. A key feature included a kill switch for rapid shutdown if needed, along with alert systems for unexpected or inappropriate content, ensuring safe and consistent operation. The tone of the chatbot was carefully calibrated to be youthful yet professional, aligning with VOXI's brand identity.

Since its launch, the chatbot significantly improved customer service outcomes. It increased the rate of query resolution without human intervention (containment rate), reduced average handling times for agent-assisted interactions, and minimized the volume of chats transferred to human agents. These enhancements delivered a faster, more engaging customer experience that better reflected VOXI’s youthful brand. Additionally, by automating routine inquiries, the chatbot freed up staff resources for more complex issues, supporting VOXI’s goal of leading the UK telecom industry in customer experience and setting a new standard in telecom chatbot innovation.