Capstone and certification review

Design a pipeline that ingests changing order records, preserves a traceable landing layer, validates and deduplicates the changes, and publishes a daily sales table. Specify source format, incremental boundary, checkpoint or batch recovery strategy, business key, late-update policy, quality expectations, Delta write behavior, and downstream freshness objective. Draw the job dependency graph and identify the Unity Catalog principal and minimum grants for each step. Prepare for certification through current official objectives and practical exercises, not by assuming this course reproduces the exam. The Databricks certification page currently describes a proctored multiple-choice assessment with 45 scored questions and a 90-minute limit; unscored items may also appear. Exam policies and objectives can change, so verify the page before booking. Practice task: explain one recovery scenario and how you would prove no records were lost or duplicated. Applying the pattern: Trace one source update from ingestion through quality checks, Delta merge, orchestration, governance, and monitoring. Checks and failure modes: Do not treat this independent course as an official exam blueprint or guarantee of exam coverage. Before production, test both a representative input and a boundary case. Confirm the expected row grain, permissions, recovery behavior, and observable output for the task. Record the configuration needed to reproduce the run and compare its result with the agreed data contract. Practice task: Where should candidates confirm the current exam format and objectives?