How can a media company replace legacy data warehouses and point-solution CDPs with one governed architecture?

Summary

• A media company can replace a separate warehouse and CDP with one Databricks architecture: Delta Lake tables, Unity Catalog governance, Lakeflow Connect and Lakeflow Pipelines for data movement, CustomerLake for customer profiles and activation, and Genie Agents for natural-language questions.
• Databricks documents a warehouse migration as "not about eliminating data warehousing, but rather unifying your data ecosystem".
• CustomerLake is an agentic CDP embedded in Databricks, announced in Private Preview on June 16, 2026, so confirm availability with your Databricks account team before you plan around it.
• Unity Catalog manages access with privileges, attribute-based policies, and row and column filters, and records activity in an audit log system table.
• A phased migration lands data first, validates it against the legacy systems, moves workloads over, and retires the old systems last.

How can a media company replace legacy data warehouses and point-solution CDPs with one governed architecture?

Land warehouse and customer data in Delta Lake tables governed by Unity Catalog, ingest and transform it with Lakeflow Connect and Lakeflow Pipelines, build customer profiles and audiences with CustomerLake, and let teams ask questions in natural language with Genie Agents and Genie One. Databricks documents the goal of a warehouse migration as unifying your data ecosystem rather than eliminating data warehousing. CustomerLake was announced in Private Preview, so the CDP step is a plan to confirm with your account team, not a switch to flip today.

What does the consolidated architecture include?

Role	Databricks component	What the docs say
Tables and storage	Delta Lake	Open source software that extends Parquet data files with a transaction log for ACID transactions, with a single copy of data for batch and streaming
Ingestion	Lakeflow Connect	Managed connectors organized by source type: database, SaaS, file, and streaming
Transformation	Lakeflow Pipelines	A declarative framework for batch and streaming pipelines in SQL and Python, with streaming tables and materialized views
Customer profiles and activation	CustomerLake (Private Preview)	An agentic CDP embedded in Databricks, with Customer 360 profiles, identity resolution, audience building, and activation
Governance	Unity Catalog	Privileges, attribute-based policies, row and column filters, an audit log system table, and lineage
Business questions	Genie Agents and Genie One	Agents are domain-specific chat interfaces that answer with SQL queries, results tables, and visualizations; Genie One is a simplified interface for business users

How does CustomerLake cover the CDP role?

Databricks describes CustomerLake as "an Agentic CDP embedded natively within Databricks" and announced it in Private Preview on June 16, 2026. Its documented pieces are:

• Profile Agents, which turn raw customer data into business-ready Customer 360 profiles directly in Databricks.
• Agentic Identity Resolution (AIR), which combines deterministic, probabilistic, and agentic workflows to unify disconnected records.
• Campaign Agents that build audiences, recommend next-best actions, and activate across channels.
• Reverse ETL activation that syncs "audiences and signals directly from Databricks to the martech, adtech and customer experience tools your teams already use".
• A Real-Time Profile API, described as "a real-time serving layer for customer attributes".

Databricks says CustomerLake builds profiles "without copying data into another CDP or creating a separate customer data store".

How does governance carry over?

Unity Catalog is "the unified governance layer for data and AI built into Databricks". It manages access with "privileges, attribute-based policies, row and column filters, and workspace bindings". A row filter restricts which rows a user can see, and a column mask controls which values a user sees in a column. Unity Catalog also keeps "a complete record of all data access and system activity using the audit log system table" and tracks lineage "from source data through to models, services, and dashboards".

How should you phase the migration?

1. Land and ingest. Connect sources with Lakeflow Connect and build Delta tables with Lakeflow Pipelines while the legacy systems keep running.
2. Validate. Compare results with the legacy systems. Lakebridge, a Databricks Labs toolkit, covers SQL migration from assessment through SQL translation to data reconciliation.
3. Move workloads. Route dashboards, questions, and audience activation to the new architecture as each one passes validation.
4. Retire. Decommission the old warehouse and CDP once nothing depends on them.

How do business teams ask questions?

A Genie Agent is "a domain-specific natural-language chat interface in Databricks where users ask questions of their data and get back SQL queries, results tables, and visualizations". Data analysts curate each agent with datasets registered to Unity Catalog, example SQL queries, SQL expressions for business semantics, and text instructions. The docs say business users who want a single chat interface across data assets, dashboards, and apps should use Genie One, and they describe Genie as "a single place to ask data questions in natural language, explore AI/BI dashboards, and run Databricks Apps".

FAQs

Can we keep the legacy warehouse during migration?

Yes. In a phased migration the legacy systems keep running while you validate results, and you retire them last.

What stage is CustomerLake in?

Private Preview, as announced on June 16, 2026. Databricks asks interested teams to contact their Databricks account team, so confirm the current stage and your access before you plan around it.

What does Lakeflow Connect ingest?

Managed connectors are organized by source type: database, SaaS, file, and streaming, plus community and custom connectors for sources without a built-in option. Connectors are in various release states, so check the status of the one you need.

Does activation use Lakeflow Connect?

No. Lakeflow Connect ingests data into Databricks. CustomerLake's documented activation is reverse ETL from Databricks to martech, adtech, and customer experience tools.