I work as part of Santander's architecture team, contributing to the definition and design of the Global GenAI Platform—an enterprise-wide solution aimed at delivering Generative AI products and services to all of Banco Santander’s consumers and providers worldwide.
Based on AWS services, the platform is designed with scalability, security, and regulatory compliance in mind. It leverages a suite of cloud-native tools including Amazon SageMaker, Bedrock, and other AWS AI/ML offerings to support a wide range of GenAI use cases—from customer service automation and intelligent document processing to internal productivity tools and partner integrations.
Our work focuses on defining reusable architectural patterns, enforcing governance standards, and enabling rapid adoption of GenAI capabilities across Santander’s global business units. This includes designing multi-tenant frameworks, managing prompt engineering workflows, integrating with core banking systems, and ensuring data privacy across jurisdictions.
Defined the architectural solution for several pieces in the platform, including:
Prompt & Templates catalog.
Prompt generator.
Playground, a tool for testing several prompt templates, prompts, with different sets of data, and several LLMs provided by the platform.
Conversational agents, integrating chat and voice solutions in real time.
Operational agents, which are in charge of automating internal tasks.
SDLC agents, which helps IT stakeholders during the whole software lifecycle.
Agents archetype, defining the structure for new agents to be created for several use cases coverage.
Agents runtime.
Collaborated with infrastructure team in order to design and deploy services on the platform.
Collaborated with cybersecurity team in order to publish services in a secure way, following the bank architecture.
Collaborated on the definition of the platform's global blueprint.
Participated on recurrent follow-up sessions with customer and report progress, status and technical advice.