Databricks Data Engineer Associate study plan

Five weeks at 6–8 hours a week. You need a Databricks workspace — this is a platform exam and the questions assume you have used Delta Lake, Lakeflow Jobs and Unity Catalog rather than read about them.

Because no section exceeds 22%, the plan is about coverage. Seven sections, none of which you can safely skip.

Week 1: the platform (6%) and ingestion (21%)

The platform section is only about three questions — cover it quickly:

• What the Databricks Intelligence Platform is and how the workspace is organised.
• The lakehouse idea: Delta Lake as the storage layer.
• Clusters and compute basics.

Then ingestion, which is 21%:

• Loading data from cloud storage into Delta tables.
• Incremental and streaming ingestion, and why reprocessing everything is the wrong default.
• Handling schema on load, including schema evolution.
• File formats and what each costs you.

Week 2: transformation and modelling (22%)

The largest section.

• Transforming data with SQL and PySpark on Databricks.
• Delta Lake operations: merge, update, delete, and time travel.
• Data modelling on the lakehouse — the bronze, silver and gold pattern, and what belongs in each layer.
• Handling duplicates, nulls and late-arriving data.
• Building idempotent transformations, so a rerun does not double-count.

Week 3: Lakeflow Jobs (16%) and CI/CD (10%)

Orchestration:

• Creating jobs, tasks and dependencies between them.
• Scheduling, and handling failures and retries.
• Passing parameters between tasks.
• Monitoring job runs and reading failures.

CI/CD:

• Git integration and repos in the workspace.
• Promoting work between development, staging and production.
• Environment-specific configuration kept out of the code.
• Automated testing of data pipelines.

Week 4: governance and security (15%), troubleshooting (10%)

Governance:

• Unity Catalog: catalogs, schemas, tables, and the permission model.
• Granting and revoking access; principle of least privilege.
• Lineage — what a table was built from and what depends on it.
• Handling sensitive data, including masking and row or column level control.

Troubleshooting, monitoring and optimisation:

• Diagnosing a failed or slow job.
• Reading the Spark UI at a basic level.
• Common performance problems: small files, skew, unnecessary shuffles.
• Optimising Delta tables, and why file layout matters.
• Cost awareness — cluster sizing and termination.

Week 5: practice and repair

• Full practice exam under real conditions: 90 minutes, 45 questions.
• Sort mistakes by section. With seven sections, patterns show up quickly.
• Rebuild the weakest, then take a second practice exam.

Where the hours go

Note how even that is. The exam is flat by design and your study should be too.

Build one pipeline

The most efficient preparation: build a small pipeline that ingests files from storage into a bronze Delta table, transforms into silver, aggregates into gold, runs as a scheduled Lakeflow Job with a dependency between tasks, is governed in Unity Catalog, and is committed to a Git repo.

Then break it. Feed it a file with a changed schema. Let a task fail and read the error. Create a thousand tiny files and watch performance degrade.

That single project touches every section, and the breaking is where the troubleshooting marks come from.

Idempotency is worth special attention. A pipeline that produces different results when rerun is a real-world failure the exam likes to describe.