In the race to build the next breakthrough in generative AI, it’s easy to overlook the foundation everything depends on: data. Poor quality, unclear lineage, and inconsistent interfaces are bottlenecks not only for production ML, but also for open collaboration. That’s where data contracts come in.

In this session, Jean-Georges “jgp” Perrin, author of Implementing Data Mesh and chair of the Linux Foundation’s Bitol project, will explore how open standards for data contracts and data products power the next wave of AI development. Using the Open Data Contract Standard (ODCS) and the Open Data Product Standard (ODPS), we’ll see how AI can finally treat data as a product, not a problem.

Expect live examples, GitHub-ready YAML, and practical insights on:
Preventing silent failures in Data & AI pipelines
Establishing trust and lineage across open datasets
Aligning data collaboration with reproducibility and compliance (without bureaucracy)

Whether you’re a developer fine-tuning models or building analytical reports, this talk will give you the tools to inject governance directly into your data and GenAI workflows — and keep moving fast without breaking things.