BBVA Cuts Customer Inquiry Response Times by More Than 15% in Italy and Germany Thanks to AI

To respond to the nearly 100,000 customer enquiries they receive every month via telephone, email and the ‘My Conversations’ messaging channel on the app and website, BBVA customer service agents in Italy and Germany previously had to consult multiple of internal documents and product and service manuals specific to each market. This was often a time-consuming task, requiring agents to search through multiple sources to find the right information for each case.

To address this need, the Digital Banks teams identified the opportunity to create specialised AI assistants that bring together all the knowledge required to answer these enquiries in a single place. Agents simply submit the customer's request in natural language, and the AI copilot turns the bank's internal knowledge into a structured response proposal in the customer's language within seconds. The agent then reviews and adapts the response before sending it. This process speeds up query resolution by reducing the average handling time for each customer interaction by more than 15%.

Used daily by more than 260 customer service agents across both countries, the tool has no access to customers’ personal or sensitive data. It is designed to help answer frequently asked questions about products, services and general procedures—for example, how to increase a transfer limit or when a newly ordered card is expected to arrive. Whenever a response depends on the customer’s specific circumstances, such as the products they hold, the conditions of their account or a particular transaction, it is always the agent who analyses the case, consults BBVA’s internal systems and provides a personalised response. In this way, AI speeds up access to information while enabling agents to devote more time to enquiries that require analysis and personalised support, serving more customers without compromising service quality.

“What makes this project particularly interesting is that the need came directly from the teams who interact with our customers every day. They identified a challenge in their daily work, and artificial intelligence enabled them to turn that idea into a solution with a tangible impact on both the business and the customer experience,” says Elena Alfaro, Head of Global AI Adoption at BBVA.

The assistants were developed using BBVA Assistants, the Group’s corporate generative AI platform, which enables employees to create and scale specialised assistants that transform business processes. Built on BBVA’s proprietary orchestration architecture, the platform combines the capabilities of OpenAI’s models with access control, knowledge management and supervision mechanisms that allow specialised assistants to be deployed for specific business needs while complying with the bank’s security and governance standards.

In June alone, customer service agents in both countries exchanged around 40,000 messages with their AI copilots while preparing responses to customer enquiries. This makes them among the most widely used assistants on BBVA Assistants, demonstrating how bringing generative AI capabilities closer to business teams enables employees to identify improvement opportunities and develop solutions with a direct impact on customer experience. This is particularly significant in BBVA's two fully digital banks. In Italy and Germany, contact centers play a fundamental role in the day-to-day relationship with customers, meaning that improving agents' responsiveness has a direct impact on customer experience.

Before going into production, the assistants were assessed under BBVA’s AI Assistant Governance Framework, an internal mandatory governance model for all AI assistants developed within the bank. The framework establishes the supervision and control measures that every assistant must meet depending on its use case, including human oversight mechanisms, reviews of the quality of the prompts and knowledge sources used, and impact assessment. This helps mitigate risks specific to generative AI, such as the use of inappropriate data or the generation of incorrect or unsupported information. In this case, the assistants incorporate technical controls that alert agents to the type of data they may use, and all outputs are reviewed before being shared with customers.

Training has also been essential to mitigating these risks. Customer service agents receive dedicated training sessions where they learn how to formulate prompts in natural language, provide context and make the most of the assistants’ capabilities. This training contributes to more accurate responses, improves the user experience for agents and accelerates adoption of the solution.