Loading…
OpenAI and Databricks at DAIS 2026: Making enterprise AI real
Margaret Amori
- Source
- Databricks
- Published
- Added to Yomu
Summary
At Data + AI Summit 2026, Databricks and OpenAI presented a partnership centered on combining OpenAI’s frontier models and agents with Databricks’ enterprise context and control. The post describes GPT models and Codex running natively on Databricks, with Unity AI Gateway governing interactions, budgets, routing, auditing, and visibility, while Agent Tools provide governed enterprise data access through MCPs. Examples include OpenAI’s marketing data foundation on Databricks, which cut storage costs by $400,000 per month, and a Hertz application built in 11 business days with GPT-5.5 and Databricks that lifted lead conversion from 60–65% to 75–80%. The sessions argue that deployment, security, evaluation, monitoring, context, and sharing comprise most of the work around enterprise agents. A joint virtual event scheduled for August will address shipping agentic applications at scale, including a Stellantis case study.
Context
The partnership is presented as a way to help enterprises move AI from pilots and research into trusted, production-scale applications. The source says the main challenge is no longer model intelligence, but the surrounding work of deployment, security, evaluation, monitoring, context, sharing, governance, and business impact.
Approach / What changed
OpenAI supplies frontier models and agents, while Databricks supplies enterprise context, data access, governance, and operational control. GPT models and Codex are available on Databricks, with Unity AI Gateway for auditing, budgets, traffic routing, and visibility, and Agent Tools for governed access to enterprise data through MCPs.
Takeaways
- Unity AI Gateway is described as governing OpenAI deployments through interaction auditing, per-user budgets, traffic routing, and security and compliance visibility.
- OpenAI rebuilt its marketing data foundation on Databricks using a Bronze-Silver-Gold medallion architecture, eliminating $400,000 per month in storage costs and enabling marketer self-service without SQL.
- Hertz used GPT-5.5 with Databricks, Lakebase, and Unity Catalog volumes to build an accountability application in 11 business days; lead conversion improved from 60–65% to 75–80%.