The Databricks Data + AI Summit 2026 signals a shift from experimentation to enterprise-scale, agentic AI. With over 30,000 attendees (representing a 36% annual increase) and global participation across more than 150 countries, Databricks is positioning itself as a foundational platform for data-intelligent applications. Last year, we wrote that Databricks went “beyond the lakehouse” by making a “bold play for the business persona.” This year, the company makes that strategy tangible by formally entering the agentic customer data platform (CDP) market with CustomerLake, an AI-native, warehouse-based platform built for enterprise-grade agentic marketing. Moreover, Databricks encourages data and AI leaders to focus on four imperatives: choice, context, cost, and control. Here are some key themes and announcements and what they mean for data and AI technology professionals.

The Four C’s: Choice, Context, Cost, And Control Of Data-Intelligent Agents And Apps

Databricks’ choice, context, cost, and control-focused model defines how to scale enterprise AI. Choice enables flexibility across proprietary and open models via the Agent Bricks platform. Context grounds AI in business semantics and meaning to improve accuracy and decision-making via Genie Ontology. Cost allows AI spend visibility and operations with Unity AI Gateway. Control unifies governance across data, models, agents, tools, skills, MCP services, and interactions in one place anchored by Unity Catalog governance extensions, Omnigent, and CustomerLake. The goal is to help organizations control access, tool invocation, and auditability while improving cost monitoring, budgeting, and request routing. Databricks now elevates the AI governance runtime control plane and extends its lakehouse with security information and event management (SIEM) capabilities. The shift is significant: Governance is no longer applied after the fact but embedded directly into how AI systems execute. This strongly positions Databricks as enterprises look to centralize oversight of AI usage while getting into the SIEM space.

Databricks Redefines The Customer Data Platform For Agentic AI

Databricks’ agentic customer data platform built on a customer data lake marks a fundamental shift from traditional CDPs that unify customer data to platforms that enable AI agents to reason, decide, and act on trusted customer intelligence. CustomerLake provides campaign and profile agents, an open partner ecosystem, and native integrations with reverse ETL to ingest, unify, and activate data across marketing and advertising technology.

Harnesses And Omnigent (Organizations Are Still Missing Steering Councils)

Omnigent from Databricks represents an emerging “meta-agent” scaffolding layer that orchestrates, governs, and executes multiple AI agents at scale. These agent harnesses combine myriad operations into a unified system, connecting data, tools, policies, and actions rather than treating agents as isolated components. This creates a cohesive system of action, allows faster scaling of AI, improves control, and reduces operational risk. In theory, it’s elegant. Meanwhile, tech executives continue to champion autonomous AI on the main stage while quietly insisting that every agent remain tethered to a human in the loop — because accountability, unlike AI, is still very much nondelegable.

Final Thoughts

Databricks’ Data + AI Summit 2026 signaled a decisive shift. Databricks is expanding beyond its core lakehouse into adjacent markets like CDP and SIEM, signaling a broader enterprise platform strategy. This creates an opportunity but also raises the bar for rigorous platform validation. Yet Databricks’ expansion into CDP and SIEM markets poses an important question: Will CMOs and CISOs abandon the platforms they have invested in, operationalized, and trusted for years in favor of a greenhorn? Databricks must prove that its unified platform delivers superior business value, especially in the SIEM market, where buyers expect day-one reliability and trust. Customers should ensure that Databricks meets requirements for each workload, particularly around latency, governance, orchestration, integration, and control, before extending beyond the lakehouse.

Share your thoughts and discuss any of these announcements with us by scheduling an inquiry or guidance session.

Share