Databricks CustomerLake Expands Campaign Agents and Profile Agents
On October 19, 2026, Databricks announced broader access to CustomerLake, its customer data platform (CDP) with AI agents. Campaign Agents entered beta, while Profile Agents became available in private preview. Databricks also reported early customer results, including 90% faster audience preparation, a potential 40% reduction in CDP total cost of ownership, and a 10-percentage-point conversion lift in one campaign.
Campaign Agents Enter Beta for Audience Creation and Campaign Activation
On October 19, 2026, Databricks moved Campaign Agents into beta and expanded access to additional enterprise customers that meet its technical requirements. The tool helps marketers build audiences, plan campaigns, and activate customer data directly from Databricks without moving that data to a separate CDP.
Campaign Agents use Databricks Genie to let marketers define audience segments in natural language. For example, a marketer might request customers who placed multiple orders during the previous six months but have not been active in the past 30 days. The system uses existing customer data and its business context to prepare the segment.
The agents are also designed to recommend next actions, personalize messages, and manage campaigns based on customer behavior. Databricks calls this approach Infinity Campaigns: rather than relying only on one-off sends, businesses can build ongoing programs that adapt to customer signals.
The functionality available in beta should not be treated as a fully autonomous marketing system. Employees still set the strategy, goals, and constraints, while AI helps automate execution.
Profile Agents Enter Private Preview to Build Unified Customer 360 Profiles
On October 19, 2026, Databricks also opened a private preview of Profile Agents. The tool is designed to combine fragmented customer data into unified Customer 360 profiles within a company’s existing Databricks infrastructure.
The update covers three core capabilities:
- Source Normalization: Profile Agents map fields from different sources into a shared structure. Customer data from a CRM, transaction system, and mobile app may use different attribute names and formats.
- Agentic Identity Resolution: The agents use deterministic, probabilistic, and LLM-based matching to connect records associated with the same customer. When creating a golden record, the system determines which attribute values to retain in the unified profile.
- Incremental Processing: Changes are processed in smaller batches so Customer 360 can be updated regularly without rebuilding the entire dataset. Data stewards can make manual corrections in ambiguous cases.
An e-commerce company, for example, could combine Shopify order history, customer support interactions, ad engagement, and predictive model outputs. The resulting profile could then support segmentation, personalization, and automated campaigns.
Databricks emphasizes that Profile Agents and Campaign Agents can be deployed independently. Companies with established customer datasets can begin using Campaign Agents without waiting for Profile Agents.
Early Customer Results From HP, Wine.com, and Philadelphia Union
On October 19, 2026, Databricks shared early results from enterprise deployments of CustomerLake. These figures describe specific implementations, not guaranteed outcomes for every business.
- HP: Prepared the data and attributes needed for marketing audiences in two hours, compared with an estimated two to three months for a traditional CDP implementation.
- Philadelphia Union: Reduced audience preparation time by 90%.
- Wine.com: Maintained 100 active audiences synchronized with Braze every day.
- Databricks Data + AI Summit: Reported a 10-percentage-point conversion lift against a control group for a personalized session-recommendation campaign.
HP: Faster Audience Preparation and Projected CDP Savings
On October 19, 2026, HP reported that the data needed to create marketing audiences was ready in two hours. The company also estimated that using its existing Databricks infrastructure could reduce its customer data platform total cost of ownership by 40%. This 40% figure is a projected saving, not an already verified reduction in operating costs.
Wine.com: Production Campaigns and Daily Braze Synchronization
On October 19, 2026, Wine.com reported that its first production email campaigns launched 23 days after the project began. The company synchronizes 100 active audiences with Braze every day, using information from 6.5 million customer profiles, purchase histories, wine preferences, and engagement data.
Philadelphia Union: 90% Faster Audience Preparation
On October 19, 2026, Philadelphia Union reported a 90% reduction in the time needed to prepare audiences. The organization uses Databricks to build segments and performs hundreds of synchronizations per month with Salesforce Marketing Cloud.
Data + AI Summit: Personalized Recommendations and Conversion Lift
On October 19, 2026, Databricks also described its own campaign for 30,000 Data + AI Summit attendees. Campaign Agents personalized session recommendations using engagement history. The company reported a 60% email open rate, a 10% click-through rate, and a 10-percentage-point increase in conversion compared with the control group.
CustomerLake Connects With Existing Marketing and Advertising Platforms
On June 16, 2026, when it initially announced CustomerLake, Databricks introduced an open integration architecture for advertising and marketing systems. Moving Campaign Agents into beta in October expanded opportunities to put that infrastructure to work.
CustomerLake uses reverse ETL to send prepared audiences and customer data to marketing activation platforms.
Databricks named Braze, Bloomreach, Iterable, Adobe, Meta, The Trade Desk, Twilio, LiveRamp, and other vendors among its partners.
An important characteristic of the architecture is that CustomerLake does not require enterprises to move all their data to a separate closed CDP. It can access existing data sources through Databricks capabilities, including Unity Catalog and Lakehouse Federation.
For example, a business can continue using Braze to deliver personalized emails and push notifications, while using Databricks to build audiences, calculate predictive metrics, and unify customer data.
CustomerLake can therefore serve as a central infrastructure for customer data management and marketing activation without replacing every tool in an existing MarTech stack.
What Databricks CustomerLake Does for Marketing Teams
Databricks CustomerLake is a customer data platform built into the Databricks Data Intelligence Platform. It brings together Customer 360 profile creation, identity resolution, segmentation, audience preparation and activation, and AI-supported marketing workflows.
CustomerLake works with data and tools already available in the Databricks ecosystem. Businesses can manage customer information within their existing data infrastructure and send prepared audiences to connected marketing and advertising systems.
For e-commerce businesses, use cases include unifying purchase history and behavioral events, creating segments, and personalizing communications. Customer service companies can combine interaction, order, and service data in a single customer profile. B2B businesses can use customer and commercial data to build segments and marketing programs according to their data models.
CustomerLake differs from standalone CDPs by running these capabilities within Databricks and using AI agents to prepare profiles and marketing audiences. At the time of the October announcement, Campaign Agents were in beta and Profile Agents were in private preview. Databricks had not announced a general availability date or public pricing.
The updates are designed to reduce manual work in customer data preparation, accelerate segmentation, and support personalized marketing campaigns using existing enterprise data sources.