BBVA develops a new technology architecture with AWS to accelerate its AI solutions
BBVA has developed a new MLOps architecture together with Amazon Web Services (AWS), integrated into ADA, the bank’s global cloud-based data and artificial intelligence platform, to accelerate the development, validation and deployment of AI models across the Group. The solution enables teams to work more autonomously, reuse common components and automate governance controls, reducing development times by up to 75% in use cases such as personalized customer recommendations and financial forecasting.
In collaboration with AWS, BBVA has implemented an MLOps (Machine Learning Operations) framework in ADA, its global analytics, data and AI platform, enabling more scalable and governed management of the AI model lifecycle. The architecture automates operational tasks and validation processes, integrating the development, testing and deployment of solutions into the bank’s technology operations. Both companies presented the solution at the annual AWS Summit event.
The architecture facilitates the work of ADA’s more than 6,500 users, including 1,000 data scientists who develop AI-based solutions at BBVA, enabling faster creation and deployment across the Group. In pilot projects such as personalized recommendations for clients or financial forecasting, the solution has reduced development times by 20 to 75 percent, and optimized infrastructure operational costs by 40 to 55 percent.
The MLOps architecture also includes governance in the development cycle for AI models. The system automates validation, traceability and control processes to enable the safe transition of models into production, keeping the bank’s review and approval mechanisms intact. Furthermore, it maintains a centralized audit trail to ensure that all machine learning models generated at BBVA comply with the security and transparency standards that apply to the financial sector. The system is essential in a sector where traceability, security and risk control are vital.
“Artificial intelligence only creates real value when it can be scaled industrially across the entire organization. The new MLOps architecture gives us a competitive advantage to accelerate the transformation of our internal operations and deliver secure and transparent AI solutions to our customers more quickly,” said Natalia Sampietro, from the Data & Analytics Enablement team at BBVA.
This transformation is based on Amazon SageMaker AI, AWS’s ecosystem of tools for building, training, deploying and managing machine learning and artificial intelligence models. One of the solution’s key advances is the creation of ephemeral development environments on AWS cloud-based machine learning infrastructure, allowing multiple teams to work, experiment and validate new functionalities in parallel without interfering with one another or affecting shared environments. Once the testing is complete, the resources are automatically eliminated, thus accelerating development cycles and optimizing infrastructure use.
“At AWS, we are very proud to collaborate with BBVA in this transformation that allows over 6,500 data professionals to accelerate the creation and deployment of AI models with autonomy and rigor. With this MLOps architecture, BBVA is demonstrating its innovative vision and its commitment to scaling AI securely and with agility on a global scale,” said Carlos Alegre Berges, Head of Sales for the FSI sector at AWS Spain.
The case study was presented during the AWS Madrid Summit, the company’s annual business event, where AWS shared its report ‘Unlocking the Potential of AI in Spain’ with more than 10,000 attendees. The report also highlights the joint project with BBVA as an example of technological transformation and advanced adoption of artificial intelligence in enterprise environments.