Databricks Data + AI Summit 2026
The largest dedicated Data + AI event in California. 800+ sessions on Mosaic AI, Unity Catalog, foundation models, and the operational reality of running AI at scale.
Presented by
Monday, June 15 – Thursday, June 18, 2026
4-day conference · Day passes available
Moscone Center
747 Howard St, San Francisco, CA 94103
20,000+ attendees expected
Data engineers · ML engineers · platform leaders
About this event
A technical room running 18 months ahead of the rest of the industry
Databricks Data + AI Summit draws roughly 20,000 attendees concentrated in Moscone. The audience is technical: data engineers, ML engineers, platform leaders, and the analytics teams behind enterprise AI.
The event's trajectory has shifted from a data warehouse conference to a full-stack AI platform event. Foundation model training, Mosaic AI, and agentic workflows now dominate the keynote, and the agenda runs 800+ sessions across data engineering, ML, generative AI, and platform tracks.
Why it matters for Robot.com
Three reasons this is an anchor event
-
Fleet data lives on Databricks
Route optimization models, delivery attribution, demand forecasting, and autonomous perception training workloads increasingly run on the Databricks platform. Attending is how teams stay current with Mosaic AI, Unity Catalog, and the serving stack.
-
Foundation model strategy is debated here
Fine-tune open models, use Databricks-hosted models, or build on OpenAI and Anthropic APIs — this is an active architectural decision for robotics teams. The Summit surfaces the patterns enterprises are actually shipping.
-
ML Ops patterns are mature here
The robotics industry is 18 to 24 months behind enterprise ML on operational discipline. Databricks talks on feature stores, model monitoring, and eval pipelines compress that gap fast.
How to work the show
A playbook for 4 days at Moscone
- Attend the Mosaic AI track. Training and fine-tuning content is where robotics perception teams find the most transferable patterns.
- Book meetings with the Databricks field team. Reference architecture conversations produce durable technical alignment.
- Go deep on one ML Ops capability per day. Evals on day one, feature store on day two, model serving on day three. One deep stack per day beats shallow surveys across all tracks.
Robot.com fit
Where this event maps to the product family
Direct fit
ML platform for R-kiwi route optimization, R-cargo warehouse orchestration, and R-top floor coverage planning. REMI remote operations analytics.
Adjacent fit
R-ads measurement and attribution data pipelines.