DataTalks x april: The Rise of Tabular ML

Our next DataTalks TLV meetup will explore one of the most exciting emerging directions in machine learning: Tabular Foundation Models.
For decades, structured data problems have been dominated by tree-based models like XGBoost and Random Forest. While deep learning transformed NLP and vision, it consistently struggled to outperform classical methods in tabular settings. Over the past year, a new paradigm has begun reshaping the field — bringing foundation model thinking into structured data.

***

## The Talk 🌐

From Trees to Transformers: The Rise of Tabular Foundation Models
Speaker: Alan Arazi
In this session, Alan will:

• Provide a clear overview of the tabular learning landscape
• Explain why deep learning historically underperformed on tabular tasks
• Introduce TabPFN — a breakthrough model trained to perform Bayesian inference over tabular datasets
• Dive into how it is trained, how it works, and what makes it fundamentally different
• Discuss what this shift means for the future of data science

This talk is designed for ML practitioners, researchers, and data professionals who want to understand what comes after gradient boosting and how foundation models may finally unlock deep learning’s potential for structured data.

***

## ‍‍About the Speaker 👨🏻‍🏫

Alan Arazi is a PhD candidate at the Technion under Prof. Roi Reichart, researching Tabular Foundation Models. After five years working with large language models at AI21, he joined Prior Labs — the company behind TabPFN — where he works at the frontier of structured data modeling.

***

## Location📍

• April Offices
• 7 Totseret HaAretz St., Tel Aviv-Yafo
• Networking and Q&A included.

We hope to see you there!