From Context to Value: Deploying AI Models on a Medallion Architecture with MCP

In this masterclass, we explored how to transform data into real value by integrating artificial intelligence models within a Medallion architecture. From a practical perspective, we saw how to structure data into layers (bronze, silver, and gold) and how these layers facilitate the deployment of scalable and well-governed AI solutions.
We introduced the Model Context Protocol (MCP) as a key element for managing model context and improving performance. The session combined theory and practice with real-world examples.

Programa

The session is structured as follows:

Introduction to the Medallion architecture
Data layers: bronze, silver, and gold
Integrating AI models into data architectures
Introduction to the Model Context Protocol (MCP)
Best practices for production deployment
Use case / live demo
Q&A

In collaboration with:

Play video Download Presentation

Taught by:

Víctor García Domínguez, Vicente Juan Magallanes

Víctor Garcia Domínguezholds a degree in Nanoscience and Nanotechnology from Universitat Autònoma de Barcelona (UAB). He completed his undergraduate thesis at Florida International University. He also holds a Master’s degree in Visual Tools to Empower Citizens from Universitat de Girona (UdG) and a Master’s degree in Data Science from Universitat Oberta de Catalunya (UOC).
He joined the Intelligent Data Science and Artificial Intelligence research group (IDEAI‑UPC), where he worked for two years and three months as a technical support specialist, collaborating closely with the group’s leadership. He is currently Head of the Artificial Intelligence team at Conversia, a role he has held for the past three years.
In this position, he leads the AI department, overseeing operations, project delivery, talent development, and the annual innovation roadmap, while also advising the executive committee on strategic corporate AI opportunities.
He has driven the deployment of generative AI solutions and delivered comprehensive corporate training programs that integrate AI expertise across the organization. He has actively participated in various conferences and technological projects, contributing to program design and organization, and delivering talks on applied artificial intelligence.

Vicente Juan Magallanes,has over nine years of experience in data analysis and Business Intelligence solutions, specializing in the design and implementation of projects focused on the Microsoft ecosystem and complementary technologies. His primary expertise lies in Power BI and Power Pivot, where he has developed dynamic dashboards and advanced visualizations that transform complex data into clear, strategic insights.
He has strong expertise in dimensional modeling (star and snowflake schemas), the creation of Tabular and Multidimensional models, and advanced use of DAX and Power Query to optimize analysis and performance. He also has extensive experience integrating Power BI with a wide range of data sources, both on‑premises and in the cloud, using platforms such as Azure and databases including SQL Server, PostgreSQL, and MySQL.
In the area of ETL processes, he has worked extensively with SQL Server Integration Services (SSIS) and Microsoft Fabric, developing robust and scalable pipelines that feed analytical models. He also has advanced SQL skills for complex querying and database optimization.
Beyond Power BI, his expertise spans the entire Microsoft Power Platform suite, including Power Apps and Power Automate, enabling him to create automated and innovative business solutions. These tools have been key to improving operational efficiency and accelerating decision‑making in the projects he has been involved in.
His ability to combine analytics, data modeling, and business automation tools positions him as a well‑rounded specialist capable of delivering strategic and sustainable solutions in dynamic, high‑demand technological environments.