Keynote – The Left Brain of AI: Why Structured Data Is AI's Next Frontier

Over the past three years, LLMs have transformed how organizations interact with unstructured data like text, images, code, and video. But the majority of valuable enterprise data is structured in millions of rows and columns of tabular data: transaction logs, ERPs, CRMs etc. It's these datasets that underpin every critical enterprise decision.

Until now enterprises have relied on antiquated ML algorithms to leverage these tabular datasets to make predictions and inform decisions. This is a slow and inaccurate process.

In this keynote, Alexandre Gerbeaux explores how Large Tabular Models (LTMs) solve this problem, and why LTMs are emerging as the next major category in AI given their promise to unlock trillions of dollars of value that have been locked in tables. He'll explain why structured data has remained one of AI's biggest blind spots, how LTMs differ from traditional machine learning and LLM architectures, and why predictive AI is entering its own foundation model era.

Through real-world examples in finance, healthcare, insurance, and manufacturing, Alexandre will show how organizations are moving from experimentation to deployment — and why this shift creates a massive new opportunity for data scientists, machine learning engineers, and forward deployed engineers.

The future of enterprise AI isn't generative — it's predictive.