BBVA and AWS launch AI-powered MLOps to slash model development time by 75%

BBVA has developed a new MLOps architecture in collaboration with Amazon Web Services (AWS). The system is part of BBVA’s global cloud-based data and AI platform, ADA. It aims to streamline the creation and deployment of AI models across the bank’s operations. The architecture integrates governance into the AI model development cycle. It automates validation, traceability, and control processes, ensuring compliance with financial sector standards. A centralised audit trail tracks all machine learning models for security and transparency.

Built on Amazon SageMaker, AWS’s ecosystem for AI and machine learning, the solution reduces development times by up to 75%. This applies to use cases such as personalised customer recommendations and financial forecasting. Over 6,500 ADA users, including 1,000 data scientists, now benefit from faster and more efficient workflows.

The project was showcased at the AWS Madrid Summit. AWS presented its report, Unlocking the Potential of AI in Spain, highlighting the joint effort with BBVA. Carlos Alegre Berges, Global Head for AWS’s relationship with BBVA, expressed pride in the collaboration. Natalia Sampietro from BBVA’s Data & Analytics Enablement team noted that the new architecture speeds up internal operations and delivers secure AI solutions to customers. The MLOps system automates operational tasks and validation, merging development, testing, and deployment into BBVA’s technology framework. It supports thousands of users and significantly cuts model development time. The architecture ensures compliance while accelerating the bank’s AI capabilities.