Databricks: Zero to Hero
Part of AI Engineering Lab · Developed by Zorost Intelligence AI Lab · https://zorost.com The most comprehensive module in the program: Databricks AI engineering from day zero to production, mirroring Zorost's own Databricks Modernization Practice (Legacy BI Migration · ETL Conversion → Lakeflow/Spark/DLT · Unity Catalog Governance · Model Serving · FinOps).
This module is the backbone of Weeks 21 to 24 of the 24-week program and stands alone as a complete Databricks learning path. Notebooks live in curriculum/week-21/ through week-24/; the files below are the concept deep-dives.
The path
| # | File | What you learn | Week |
|---|---|---|---|
| 00 | 00-day-zero-setup.md | Account, workspace, CLI auth, first notebook | 21 |
| 01 | 01-unity-catalog.md | Metastore, catalogs, schemas, volumes, grants, lineage | 21 |
| 02 | 02-compute.md | Clusters, SQL warehouses, serverless, DBU pricing | 21 |
| 03 | 03-delta-lake.md | ACID, time travel, VACUUM/OPTIMIZE, liquid clustering, medallion | 21 |
| 04 | 04-dbsql.md | Databricks SQL, SQL editor, AI/BI dashboards | 21 |
| 05 | 05-pyspark.md | PySpark DataFrames, performance, Photon | 22 |
| 06 | 06-pipelines-jobs.md | Lakeflow Pipelines (DLT), streaming, Lakeflow Jobs, Connect | 22 |
| 07 | 07-mlflow-experiments.md | MLflow tracking, model registry, autologging | 23 |
| 08 | 08-feature-engineering.md | UC feature tables, FeatureLookup, point-in-time joins | 23 |
| 09 | 09-model-training.md | Classic ML, AutoML, Databricks Model Training | 23 |
| 10 | 10-model-serving.md | Serving endpoints, AI Gateway, external models | 23 |
| 11 | 11-vector-search-rag.md | Vector Search indexes, Delta Sync, RAG patterns | 23 |
| 12 | 12-ai-functions-genie.md | AI functions (ai_query…), Genie spaces | 23 |
| 13 | 13-agents.md | Agent Framework (Mosaic AI), Agent Bricks, Agent Evaluation | 23 to 24 |
| 14 | 14-apps-dashboards.md | AI/BI dashboards, Databricks Apps | 24 |
| 15 | 15-dabs-ci-cd.md | Asset Bundles, CLI, Git folders, CI/CD | 24 |
| 16 | 16-governance-security.md | Row filters, column masks, dynamic views, audit | 24 |
| 17 | 17-finopps-cost.md | Billing system tables, cost dashboards, FinOps | 24 |
| 18 | 18-certification-path.md | Databricks certifications and the hero's recap | 24 |
The capstone
capstone/, ZoroLogistics Lakehouse Intelligence: the full stack from Weeks 21 to 23 (medallion lakehouse, streaming pipeline, point-in-time ETA model, Vector Search RAG, Genie space) shipped as a Databricks Asset Bundle with CI/CD, governance, and a FinOps dashboard.
How to use this module
- Weeks 21 to 24 give you the schedule: follow the weekly checklists in the Excel tracker; each week links here for concept depth.
- Upload, don't rewrite: the weekly notebooks run in any Databricks workspace (workspace files → Repos), or run the SQL parts in the SQL editor.
- Everything here is verified against the official Databricks documentation (docs.databricks.com), cited in each file's Sources section. Databricks renames features at a brisk pace; each file notes "verify against live docs" where it matters.
- Free tier reality: Databricks offers a free trial with serverless SQL and notebooks; keep the cost notes in file 17 in mind from day one.
Related
- Knowledge base: Databricks overview
- Week 21: Day Zero · Week 22 · Week 23 · Week 24
- Zorost Databricks Modernization Practice
© 2026 Zorost Intelligence LLC · https://zorost.com