Adstra joins Databricks CustomerLake as identity resolution partner

Adstra has been named a launch partner for CustomerLake, Databricks' newly announced agentic Customer Data Platform (CDP), with the company's Conexa identity resolution platform embedded directly within the governed data environment. The integration, announced at the Databricks Data + AI Summit in New York, is available immediately.

The arrangement means that enterprise marketing and data teams can resolve customer identities across offline and online signals without extracting records from the environments in which they already sit. Adstra says its Conexa platform enriches first-party profiles with third-party signals while keeping data inside existing privacy and compliance controls, avoiding the data-movement risks that have long complicated conventional CDP deployments.

What CustomerLake does

CustomerLake is positioned by Databricks as a next-generation CDP built natively on its Lakehouse architecture. Rather than operating as a separate layer that ingests copies of customer data, it is designed to work directly against the governed data foundation, exposing agentic interfaces that marketing teams can use to unify data, automate campaign decisions and personalise customer experiences at scale.

Tasso Argyros, VP Engineering at Databricks, said that existing CDPs "helped establish the importance of unified customer data, but the AI era requires a new approach" in which customer intelligence is assembled inside the governed data layer rather than moved into a separate system. The platform supports an open partner ecosystem spanning identity, enrichment, activation, measurement and customer experience, a structure Databricks says is intended to prevent new vendor silos from forming as enterprises scale their AI-driven marketing stacks.

Rick Erwin, chief executive of Adstra, said most enterprises still face a fundamental problem: identity layers that cannot accurately connect a given individual across touchpoints, or that operate as black boxes with limited auditability. "CustomerLake changes that by building customer intelligence directly where the data already lives," he said. "With Conexa embedded natively as an identity layer, marketers and AI agents can operate from the same trusted customer understanding."

Market context and competitive landscape

The CDP market has been under structural pressure for several years, with critics arguing that standalone CDPs duplicate data unnecessarily and create governance headaches as privacy regulation tightens. A composable or "headless" CDP approach, where identity and activation capabilities are embedded into an existing data platform rather than run in parallel, has gained traction among enterprise data teams seeking to reduce stack complexity.

Databricks competes in the broader data-platform space against Snowflake, which has its own partner ecosystem and native application framework. Both vendors are pushing to make their Lakehouse and Data Cloud offerings the preferred substrate for AI-driven marketing workloads, a segment that analyst houses project will grow significantly as brands invest in first-party data infrastructure following the deprecation of third-party cookies across major browsers.

For Adstra, the CustomerLake partnership extends its reach into the Databricks customer base without requiring joint customers to adopt a separate SaaS CDP. The company did not disclose commercial terms, revenue-sharing arrangements or the number of joint customers currently using the integrated stack.

Regulatory and privacy read-across

The emphasis on in-environment enrichment is partly a regulatory response. Under GDPR, the UK Data Protection Act and the patchwork of US state privacy laws, moving personal data between systems increases compliance surface area and data-transfer risk. Building identity resolution natively into a governed Lakehouse environment may simplify data-processing agreements and reduce the number of sub-processors that enterprise legal teams need to manage.

CustomerLake's composable architecture could also ease compliance with the EU AI Act's transparency obligations for automated decision-making, given that marketing personalisation driven by AI agents is likely to fall within scope as the Act's provisions phase in through 2026 and 2027. How Databricks and Adstra structure model cards, audit logs and explainability outputs for those agents will be a near-term test of whether the "governed" positioning holds under regulatory scrutiny.