BBVA's AWS-Powered MLOps Architecture Cuts AI Development Time by 75%

BBVA has developed a new MLOps architecture with Amazon Web Services (AWS) for its global cloud-based data and AI platform, ADA. The system aims to speed up the creation and deployment of AI solutions while ensuring compliance with financial sector standards. It was showcased at the AWS Madrid Summit as part of a report on AI adoption in Spain. The architecture automates operational tasks and validation processes, integrating development, testing, and deployment into BBVA’s technology operations. It also includes governance in the AI model development cycle, ensuring traceability, control, and compliance through a centralised audit trail.

Built on Amazon SageMaker, the solution reduces development times for AI models by up to 75% in areas like personalised customer recommendations and financial forecasting. It supports over 6,500 ADA users, including 1,000 data scientists, by creating temporary development environments on AWS cloud infrastructure. This allows multiple teams to work in parallel without disrupting shared systems.

Natalia Sampietro, from BBVA’s Data & Analytics Enablement team, explained that the new system accelerates internal transformation. It delivers secure and transparent AI solutions to customers more efficiently. The MLOps architecture strengthens BBVA’s ability to deploy AI models quickly and securely. It improves collaboration across teams while meeting strict regulatory requirements. The project demonstrates how financial institutions can adopt advanced AI tools at scale.