CERTIFIED-DATA-ANALYST-ASSOCIATE Question #20: Real Exam Question with Answer & Explanation

Question

A data team has been given a series of projects by a consultant that need to be implemented in the Databricks Lakehouse Platform. Which of the following projects should be completed in Databricks SQL?

Options

• ATesting the quality of data as it is imported from a source
• BTracking usage of feature variables for machine learning projects
• CCombining two data sources into a single, comprehensive dataset
• DSegmenting customers into like groups using a clustering algorithm
• EAutomating complex notebook-based workflows with multiple tasks

Explanation

Combining two data sources into a single, comprehensive dataset (C) is the correct answer because this is a classic SQL operation - joining or unioning tables - which is exactly what Databricks SQL is designed to handle: querying, transforming, and merging structured data using SQL syntax in a governed, interactive environment.

Why the distractors are wrong:

• A (data quality testing on import) is better suited to Delta Live Tables or a data engineering pipeline with expectations/constraints, not SQL's interactive query layer.
• B (tracking feature variable usage for ML) belongs in the Feature Store, a dedicated MLflow-integrated component for managing ML features.
• D (customer clustering with an algorithm) is a machine learning task requiring Python/Spark ML libraries - best done in a Databricks Notebook with MLlib or scikit-learn.
• E (automating multi-task notebook workflows) describes Databricks Workflows (Jobs), the orchestration tool for pipeline automation.

Memory tip: Think of Databricks SQL as the "analyst's workspace" - if a business analyst could reasonably do it in a SQL editor (querying, joining, aggregating), it belongs in Databricks SQL. Anything requiring code, ML models, or workflow orchestration lives elsewhere in the platform.

No community discussion yet for this question.