A. Correct.
Data engineers primarily build, manage, and optimize pipelines using engineering services, while Databricks SQL may be used secondarily for querying and analysis.

B. Incorrect.
Business analysts are typical primary users of Databricks SQL because they explore data, write SQL, and consume dashboards directly in the SQL workspace.

C. Incorrect.
A SQL analyst's main workflow is SQL querying and analysis, so Databricks SQL is a primary service for this role rather than a secondary one.

D. Incorrect.
A business intelligence analyst primarily uses SQL warehouses, dashboards, visualizations, and BI integrations, all of which make Databricks SQL a primary interface.

E. Incorrect.
A data analyst commonly uses Databricks SQL as the main place to query tables, build visualizations, and share dashboard results.

A. Incorrect.
A table being larger than 10 GB does not decide whether DROP TABLE removes files. File deletion depends on whether the table is managed or external.

B. Incorrect.
A missing visible location is not the reason. External tables store data outside the managed storage location, so DROP TABLE removes metadata but leaves files.

C. Incorrect.
A table being smaller than 10 GB also does not control DROP TABLE behavior. External table files remain because Databricks only drops the table metadata.

D. Incorrect.
For a managed table, dropping the table normally removes both table metadata and the managed data files. The files remaining indicates an external table.

E. Correct.
Dropping an external table removes the metastore metadata but does not delete data stored at the external location. Managed table data is deleted when the table is dropped.

A. Incorrect.
reduce() aggregates array elements into a single result. It does not expand one row into multiple rows.

B. Incorrect.
array() constructs an array value. It does not explode an existing array column into multiple output rows.

C. Incorrect.
flatten() collapses nested arrays into a single array. It still returns an array rather than one row per element.

D. Correct.
explode expands an array or map into multiple rows and duplicates the other selected column values for each generated row.

E. Incorrect.
array_distinct() removes duplicate elements inside an array, but it keeps the result as one array value in the same row. The requirement is to create one output row per array element, which is what explode() does.

A. Incorrect.
Visualizations are created from queries in Databricks SQL, but Data Explorer is specifically for browsing data objects, metadata, sample data, and permissions.

B. Incorrect.
Third-party BI connectivity is handled through Databricks SQL warehouses and connectors, not Data Explorer itself.

C. Correct.
Data Explorer is used to discover and manage data objects, inspect metadata and sample data, and manage permissions where the user has the required privileges.

D. Incorrect.
Data Explorer is for discovery, metadata, previews, and permissions. It is not primarily the feature used to run arbitrary UPDATE statements.

E. Incorrect.
Dashboards support exploration and sharing, but they are not the Data Explorer feature. Data Explorer focuses on object discovery, metadata, previews, and access control.

A. Correct.
LEFT SEMI JOIN returns rows from the left table that have a match on the right. LEFT ANTI JOIN returns rows from the left table that do not have a match.

B. Incorrect.
That describes an outer-style result with right-table order columns. LEFT SEMI and LEFT ANTI joins only return rows from the left table, filtered by whether a match exists.

C. Incorrect.
LEFT SEMI and LEFT ANTI are opposites: one keeps matching left rows and the other keeps non-matching left rows.

D. Incorrect.
Databricks SQL supports LEFT SEMI JOIN and LEFT ANTI JOIN, so the statements do not fail for that reason.

E. Incorrect.
LEFT SEMI JOIN does not project any columns from the right table, even customer_id. It simply keeps left-side customers that have a matching order.

A. Incorrect.
Databricks SQL is useful for in-platform querying and dashboards, but it is not a complete replacement for every feature in Tableau, Power BI, or Looker.

B. Incorrect.
Calling it a lower-functionality substitute is too broad. Databricks SQL has its own SQL warehouse, query, dashboard, and governance strengths while still integrating with BI tools.

C. Incorrect.
It is not an exact substitute because external BI tools have their own semantic layers, presentation workflows, and enterprise reporting features.

D. Incorrect.
Databricks SQL dashboards support quick analysis and sharing, but dedicated BI tools are still commonly used for polished enterprise reporting workflows.

E. Correct.
Databricks SQL complements external BI tools by providing in-platform querying, visualizations, dashboards, and SQL warehouse access while still integrating with tools like Tableau and Power BI.

A. Incorrect.
Notebooks can run code and create visualizations, but the question specifically asks for serverless SQL compute and dashboard-oriented SQL analysis.

B. Incorrect.
Tableau is an external BI tool. The requirement asks for a Databricks Lakehouse Platform service or capability.

C. Incorrect.
Databricks Machine Learning is for ML workflows, not the SQL query, serverless warehouse, and dashboard workflow described here.

D. Correct.
Databricks SQL provides SQL warehouses, including serverless options, and supports SQL queries, visualizations, and dashboards in the Databricks platform.

E. Incorrect.
Delta Lake is the storage and table format layer. It does not by itself provide the SQL UI, serverless warehouse, visualization, and dashboard experience.

A. Incorrect.
A model serving endpoint is not the best answer because it does not match the Databricks/Spark behavior being tested. The key point is: Dashboard schedules refresh dashboard queries at configured times so business users see updated results without manually running every query.

B. Correct.
Dashboard schedules refresh dashboard queries at configured times so business users see updated results without manually running every query.

C. Incorrect.
A cluster init script is not the best answer because it does not match the Databricks/Spark behavior being tested. The key point is: Dashboard schedules refresh dashboard queries at configured times so business users see updated results without manually running every query.

D. Incorrect.
A Delta VACUUM command is not the best answer because it does not match the Databricks/Spark behavior being tested. The key point is: Dashboard schedules refresh dashboard queries at configured times so business users see updated results without manually running every query.

A. Correct.
Databricks SQL alerts evaluate query results and can notify users when a defined condition is met.

B. Incorrect.
A notebook markdown cell is not the best answer because it does not match the Databricks/Spark behavior being tested. The key point is: Databricks SQL alerts evaluate query results and can notify users when a defined condition is met.

C. Incorrect.
A table comment is not the best answer because it does not match the Databricks/Spark behavior being tested. The key point is: Databricks SQL alerts evaluate query results and can notify users when a defined condition is met.

D. Incorrect.
A Unity Catalog volume is not the best answer because it does not match the Databricks/Spark behavior being tested. The key point is: Databricks SQL alerts evaluate query results and can notify users when a defined condition is met.

A. Incorrect.
DROP TABLE is not the best answer because it does not match the Databricks/Spark behavior being tested. The key point is: Parameters and dashboard filters let viewers adjust query inputs such as region while preserving the shared dashboard structure.

B. Incorrect.
A warehouse restart is not the best answer because it does not match the Databricks/Spark behavior being tested. The key point is: Parameters and dashboard filters let viewers adjust query inputs such as region while preserving the shared dashboard structure.

C. Incorrect.
A file upload widget is not the best answer because it does not match the Databricks/Spark behavior being tested. The key point is: Parameters and dashboard filters let viewers adjust query inputs such as region while preserving the shared dashboard structure.

D. Correct.
Parameters and dashboard filters let viewers adjust query inputs such as region while preserving the shared dashboard structure.