Ole Olesen-Bagneux - auteur O’Reilly et Chief Evangelist chez Actian - explore les défis de la gestion des métadonnées en entreprise, mais surtout le formidable potentiel qu’elles représentent pour l’intelligence artificielle.

La réalité du metadata management est rarement abordée en profondeur dans la littérature technique, ni même dans les recommandations des éditeurs de technologies : on cherche avant tout à expliquer comment fonctionne la technologie.

Or, cette approche fait abstraction du contexte organisationnel, nuisant à l’implémentation de ces technologies. Dans le cas des métadonnées, cela limite leur potentiel et empêche l’interopérabilité entre les nombreux répertoires généralement présents dans les entreprises.

Pourtant, une perspective nouvelle s’ouvre : considérer ces répertoires de manière holistique, comme une stack technologique. Une approche qui offre une vision plus fine et plus fidèle de l’écosystème IT, et qui permet la création de nouveaux métamodèles via des knowledge graphs, fondés sur les cartographies existantes du système d’information.

Lors de cette session, découvrez :

• La réalité de la gestion des métadonnées en entreprise

• Comment offrir à l’IA un contexte riche et robuste

• L’apport de l’IA agentique

• Les ontologies pour le Model Context Protocol (MCP)

O'Reilly Author and Chief Evangelist, PhD Ole Olesen-Bagneux takes a deep dive into the challenges of metadata management in enterprises, and the great potential metadata represents for Artificial Intelligence. 

The reality of metadata management is – crucially – not properly addressed in most tech literature, as well as in the guidance from technology vendors. This is not a result of suspicious intentions, but a natural outcome of what is sought communicated: How technology works. 

However, this leaves out the enterprise context, and accordingly implementations of technologies suffer. For metadata, this is a problem that limits the potential and interplay of the many metadata repositories normally found in an enterprise. 

A great perspective unfolds if we consider metadata repositories more holistically as a stack, giving improved perception of what the IT landscape of an enterprise is truly like. 

Furthermore, this approach solidifies how to craft new metamodels in knowledge graphs, because they meticulously consider the existing mappings of the IT landscape. 

This is the key to unparalleled solid context for Artificial intelligence. 

Discover:

*The reality of metadata management in enterprises

* Agentic AI for the enterprise

* Ontologies for Model Context Protocol (MCP)

In this episode, I sit down with Ole to discuss his new book, "Fundamentals of Metadata Management." We move past the simple definition of "data about data" to a more nuanced view of metadata as something that exists in two places at once , serving as a pointer to find information elsewhere. Ole introduces his core concept of the "MetaGrid"—the interconnected, yet siloed, web of metadata repositories that already exists within every large organization across various teams and technologies. He argues that the key to better metadata management is not to build a new monolithic system but to recognize, document, and integrate the MetaGrid that's already there, hiding in plain sight. The conversation also covers the impact of the AI hype cycle , the lessons learned from the Data Mesh movement , the sociological incentives that help or hinder metadata projects , and the cultural clash between the worlds of data engineering and library science.

Whether it's to adhere to regulations, access markets by meeting specific standards, or devise data analytics and AI strategies, companies today are busy implementing metadata repositories—metadata tools about the IT, data, information, and knowledge in your company. Until now, most of these repositories have been implemented in isolation from one another, but that practice lies at the core of problems with data management in many companies today. Author Ole Olesen-Bagneux, chief evangelist at Actian, shows you how to masterfully manage your metadata repositories by properly coordinating them. That requires a data discovery team to increase insights for all key players in enterprise data management, from the CIO and CDO to enterprise and data architects. Coordinating these repositories will help you and your organization democratize data and excel at data management. This book shows you how. Learn what metadata repositories are and what they do Explore which data to represent in these repositories Set up a data discovery team to make data searchable Learn how to manage and coordinate repositories in a meta grid Increase innovation by setting up a functional data marketplace Make information security and data protection more robust Gain a deeper understanding of your company IT landscape Activate real enterprise architecture based on evidence

What if the world's metadata were interconnected in a decentralized way? This is part of the vision of Ole Olesen-Bagneux's Meta Grid. We have a chat about what it is, and its implications for the data industry.