Customer Service Automation · Enterprise Knowledge Search
BBVA wanted to make digital banking interactions more natural and personalized for customers, while also giving branch employees faster access to current product and service information so they could handle requests more efficiently.
Reported impact
BBVA is rolling out a more personalized banking assistant that can answer questions and complete common transactions in natural language, while its employee copilot gives agents faster access to up-to-date information so they spend less time on low-value tasks and more time serving customers.
BBVA upgraded its in-app virtual assistant Blue and launched a separate AI copilot for branch employees as part of a production generative AI program in Spain. Blue is embedded in the BBVA app as the serving layer for customer interactions and uses large language models (LLM) for intent interpretation and language generation. Within each active session, the assistant uses conversational context, previous interactions, and the customer’s normal transactions to answer questions and execute common account and card operations in natural language. BBVA constrained the assistant to the banking domain and added validations and guardrails so responses stay aligned with commercial, reputational, and legal requirements. Customer data remains inside BBVA environments and is used only in the live conversation. For employees, BBVA built an AI-based assistant over a daily updated catalog of more than 30,000 product and service references so agents can retrieve current information quickly during customer service work.
Customers interact with Blue in the BBVA app, where LLMs interpret the request using conversation context, prior interactions, and normal transaction patterns, then return tailored financial information or execute eligible account and card actions. For employees, the AI-based assistant retrieves answers from a daily updated catalog of BBVA product and service references for use during customer support.
"Blue now has enhanced abilities to interact with customers using natural language, provide tailored information on their finances, and perform some of the most common account and card transactions."
"The assistant is based on more than 30,000 references, which encompass an entire catalog of BBVA products and services that are updated on a daily basis."
"The new Blue kicks off with up to 150 queries and transactions available for customers, which translates into answering over 3,000 questions."
"The private information of each user is only used in the context of the active conversation, and is not stored or used outside of BBVA environments."
"This new system is currently available to more than 100 agents"
"the virtual assistant improves user experience thanks to its large language models (LLM) interpretation and language generation capabilities."
"validations and guardrails were put in place that guarantee that all responses are aligned with the bank’s commercial, reputational and legal requirements"
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
In-house Sample Apr 8, 2025 Bank of America Bank of America needed to improve employee productivity and client service at scale across a global workforce of 213,000 people, while reducing routine support demand, accelerating employee access to information,… 20 million clients now actively using the virtual assistant Erica · Erica for Employees · PRIVATE BANK In-house Sample Feb 7, 2025 Allianz Customer care agents and operational experts across Allianz entities in Germany had to navigate the AWISO knowledge library, an extensive repository of thousands of documents, making information retrieval slow and… accuracy rate exceeding 90% for common queries AWISO knowledge library · Enterprise Knowledge Assistant (EKA) · Future Cloud platform In-house Sample May 8, 2024 Mapfre Mapfre wanted to improve customer experience and streamline insurance operations such as claims handling, risk assessment, underwriting support, document management, internal knowledge access, and digital content… 75 use cases under study for the application of generative AI ControlExpert · Cyberwrite · Holistic AI 