Databricks ended weeks of speculation today by unveiling CustomerLake, an agentic CDP. The news was shared at the company’s Data + AI Summit in San Francisco and signals Databricks’ formal entry into the martech space. Earlier, in March, Databricks moved into the security arena with its Lakewatch product. CustomerLake equips marketers and data teams with a fleet of agents that continuously analyze behavior, make decisions, and take action. According to the company, this agentic workforce can power always-on, personalized customer experiences up to 1 billion times per day. Built on Databricks’ lakehouse architecture and governed through Unity Catalog, CustomerLake unifies customer data, identity resolution, audience creation, campaign automation, and activation. With CustomerLake, Databricks is targeting marketing teams that are heading into a future where agents will be used internally by marketers, while those same marketers will also need to engage with agents deployed by customers to research and compare products. Databricks argues that most legacy martech tools were not designed for either of these use cases. CustomerLake, image courtesy of Databricks. Does the agentic era demand a new kind of CDP? Traditional CDPs typically use a waterfall approach: campaigns are orchestrated and launched across many disconnected tools, can take weeks to go live, and keep customer data isolated from the company’s primary AI platform. This leads to fragmented identities that make true, large-scale personalization unachievable. The agentic era, by contrast, calls for real-time context, data access, and execution. With CustomerLake, Databricks is bringing the CDP natively into the…