DP-700 - Ingest and Transform Data - Section 2.5

Choose between Dataflows Gen2, notebooks, KQL, and T-SQL for transforming batch data.

Choose among Dataflows Gen2 (Power Query), PySpark notebooks, T-SQL in a Warehouse, and KQL for transforming batch data. Weigh code-first flexibility against low-code maintainability, and recognise that KQL is best suited to log and telemetry data rather than relational batch workloads.

Dataflows Gen2 (Power Query) versus PySpark notebooksT-SQL in a WarehouseKQL for log and telemetry datacode-first versus low-code transformationskill set and maintainability trade-offs

Practice question for this objective

Free sampleIngest and Transform Datahard

A finance team in Microsoft Fabric must clean a daily CSV extract by trimming whitespace, splitting a full-name column, replacing error values, and merging a small lookup before landing the result in a Lakehouse table. The people who will own and edit this transformation are business analysts comfortable with Excel but not with Python or SQL, and the team wants them to maintain the logic visually without writing code. Which transformation engine best fits this team and workload?

  • AA Dataflow Gen2 built in Power Query that applies the trims, splits, replacements, and merge steps visually and lands the result in the Lakehouse. Correct
  • BA PySpark notebook that reads the CSV into a DataFrame and applies the trims, splits, replacements, and join before writing the Lakehouse table.
  • CA T-SQL stored procedure in a Warehouse that uses string functions and a JOIN to shape the data and INSERT it into the target table.
  • DA KQL query over an Eventhouse that parses and reshapes the rows before exporting the cleaned output back to the Lakehouse table.
Choose Dataflows Gen2 with Power Query when the owning team needs a visual low-code experience to shape batch tabular data without writing code. The deciding factor is who maintains the logic: Power Query in Dataflows Gen2 records each cleansing operation as a visual applied step, so Excel-literate analysts can trim, split, replace, and merge and revisit the work later without Python, SQL, or KQL, which the other engines all require.

Why A is correct: Dataflows Gen2 expose Power Query, a visual low-code mashup tool whose applied-steps list lets Excel-literate analysts trim, split, replace, and merge without code, matching both the workload and the owning team's skill set.

Why B is wrong: A PySpark notebook can do every step, but it requires Python skills the analysts do not have, so day-to-day maintenance would fall outside the owning team and break the stated visual, code-free requirement.

Why C is wrong: T-SQL handles the transformation logic well, yet it demands SQL fluency the analysts lack; the requirement is explicitly a visual, no-code maintenance experience, which a hand-written procedure does not provide.

Why D is wrong: KQL targets log and telemetry analytics, not visual file cleansing, and it is another code language unfamiliar to the analysts, so it fits neither the relational CSV workload nor the no-code maintenance goal.

See more DP-700 practice questions, answers explained.

More in this domain

Back to all Ingest and Transform Data objectives, or the DP-700 cert hub.

Examworthy is not affiliated with or endorsed by Microsoft. Original, blueprint-aligned practice material only.