Two volumes. 37 chapters. ~1600 pages of practitioner depth.

1. Part I (Ch 22–24)

   Platform Foundations

   From OSS Spark to the Databricks Data Intelligence Platform. Workspaces, notebooks, Git folders, the Assistant. Compute: classic, serverless, SQL warehouses, AI Runtime, GPUs.

2. Part II (Ch 25–30)

   Unity Catalog & Governance

   UC architecture, the three-level namespace, volumes. Access control with ABAC and governed tags. Identity, service principals, OAuth, secret scopes. Managed tables with Delta + Iceberg/UniForm. Liquid Clustering + Predictive Optimization. System tables and HMS → UC migration.

3. Part III (Ch 31–36)

   Data Engineering at Scale

   Ingestion with Auto Loader and Lakeflow Connect. Lakeflow SDP (Spark Declarative Pipelines). Lakeflow Jobs orchestration. Declarative Automation Bundles, CLI, SDK, Terraform. CI/CD with GitHub Actions + OIDC. Performance tuning with Photon.

4. Part IV (Ch 37)

   Bridge into Volume 4

   UC Metrics and the data-product mindset that makes the AI layer in Volume 4 trustworthy.

1. Part V (Ch 38–42)

   Analytics & Natural-Language BI

   Databricks SQL in production. External BI (Tableau, Power BI, Looker, dbt). AI/BI Dashboards. AI/BI Genie. The AI SQL function family, ai_query, ai_parse_document, ai_classify, ai_extract, ai_gen, ai_forecast, vector_search.

2. Part VI (Ch 43–46)

   Mosaic AI: Models & Retrieval

   Model Serving and the AI Gateway. Foundation Model APIs and external models. Vector Search and RAG patterns. MLflow 3, experiments, UC Model Registry, traces, evaluation.

3. Part VII (Ch 47–49)

   The ML Lifecycle

   Feature Store on Unity Catalog. MLOps, promotion as alias movement, champion/challenger, traffic splitting. Lakehouse Monitoring for data and model drift.

4. Part VIII (Ch 50)

   Distributed Training

   TorchDistributor, DeepSpeed, Ray on Databricks, and serverless GPU for fine-tuning.

5. Part IX (Ch 51–52)

   Agentic AI

   Agent Bricks, classification and information extraction. The Multi-Agent Supervisor with MCP, agent evaluation, and deployment.

6. Part X (Ch 53–54)

   Operational Data + Capstone

   Lakebase, managed Postgres in the lakehouse. End-to-end retail-intelligence capstone that pulls every V3 and V4 piece together.

7. Part XI (Ch 55)

   Close

   Certification paths (Data Engineer Associate/Professional, ML Associate/Professional) and what comes next on the platform.

Free sample topics

Long-form practitioner previews of the book's content are published as free topic guides: Unity Catalog, Lakeflow SDP, Asset Bundles, and CI/CD on Databricks from Volume 3. RAG on Databricks, Agent Bricks, Multi-Agent Systems, and MLflow 3 from Volume 4.