


AI x Relational Workshop by Kumo
Train your first graph-transformer-based model, say farewell to feature engineering, and ship useful AI applications using data in your warehouse.
Join us in SF for an evening exploring how new AI models based on graph transformers are completely changing predictive modeling, recommender systems, and even RAG applications. (Consider this the farewell party for feature engineering.)
It’s a combination of tech talks, Q&A, examples and lessons learned from real-world applications, hands-on practice, and a community mixer. We’ll cover breakthrough approaches that let you train models directly on multi-table schemas without the usual feature engineering headaches.
You'll see how AI models based on Graph Transformer architectures open new possibilities for both predictive applications and entity embedding generation at scale.
And you’ll meet other data scientists, ML engineers, and AI builders who are actually shipping AI applications.
Featured Speakers
Jure Leskovec - Co-Founder & Chief Scientist @ Kumo, Stanford Professor, Former Chief Scientist @ Pinterest
Ben Shahshahani - Chief AI Officer @ Cleveland Clinic
Effy Fang - Data Scientist @ Kumo
Event Format
4:30-5:00 PM: Networking
Meet peers and speakers over food and drinks
5:00-6:00 PM: Technical Sessions
Graph Transformer architecture fundamentals
Training your first graph-transformer-based AI model for relational data (demo)
Real-world applications, fireside chat with the Chief AI Scientist of Cleveland Clinic
6:15-7:00 PM: Mixer & Hands-on Workshop
(Optional) Get practical experience with sample datasets, guided by Kumo’s ML engineers and data scientists
Chat with your peers over free food and drinks
Who Should Attend
Data Scientists & ML Engineers
You'll learn how to:
Complete modeling projects up to 20x faster
Eliminate complex data preparation steps
Build models that understand entity relationships
Expand predictive capabilities across business domains
AI Engineers
You'll learn ways to:
Generate accurate embeddings from relational data
Enhance RAG systems with structured data representations
Create more intelligent interfaces to data warehouses
Registration
Registration is free. Approval is required. Applications will be reviewed to ensure a relevant technical audience. Limited to 50 participants to ensure meaningful discussions and personalized guidance.
Security won't let you enter without an approved registration.
