Data preparation with DataBrew and SageMaker: DEA-C01 study notes and practice questions
What to know about Data preparation with DataBrew and SageMaker for the DEA-C01 exam, from the Data Operations and Support domain (22% of the exam guide), then 1 free question to test it. Part of the DEA-C01 study guide.
Data prep questions usually describe users who don’t write code, or need a quick visual way to clean data, and ask which tool fits.
Tools
Tool
Best for
AWS Glue DataBrew
Visual, no-code cleaning and normalizing with 250+ built-in transformations; saves steps as a reusable recipe that runs as a scheduled job; profile jobs show statistics, missing values and outliers; detects and masks PII
SageMaker Data Wrangler (in SageMaker Canvas)
Preparing features for machine learning, with data quality and insights reports, then exporting to pipelines
Glue Studio visual ETL
Drag-and-drop Spark jobs, with generated code you can edit
Glue jobs (Spark, Python)
Complex logic, large scale, full control
Athena SQL / CTAS
SQL-skilled users reshaping data in place
Amazon Q data integration in Glue
Generates Glue jobs from natural language
Typical preparation steps
Remove duplicates and handle missing values (drop, fill or flag).
A public library's analysts receive patron-visit exports from 30 branches, and the files have inconsistent date formats, duplicate barcodes and blank branch codes. The analysts know SQL but do not write Python, and the cleaned data must refresh every Monday before the weekly report.
The data team wants the cleaning steps to be visible and reusable, and it wants the cleaned output written to S3 without manual runs.
Which approach meets these requirements?
Choose one answer.
AA. Ask an engineer to write a Python notebook each week that cleans the files, and have the analysts run the notebook by hand on Monday.
BB. Write a set of Athena views that clean the files on each query, and have the report read the views directly.
CC. Build the cleaning steps as a DataBrew recipe, and schedule a recipe job that writes the output to S3 every Monday.
DD. Load the raw files into Amazon Quick Sight and fix the duplicates and blank codes in its dashboard fields before each report.
Show the answer and explanation
Answer: C
A. Wrong: A weekly notebook depends on an engineer and a manual run, and the analysts do not write Python. The option misses the requirement that the refresh happen without manual work.
B. Wrong: Views clean the data on every query, which repeats the work and makes the cleaning steps hard to inspect. The option does not produce a cleaned output in S3 for the report.
C. Right: Correct. DataBrew lets analysts build the cleaning steps visually, saves them as a recipe, and runs them as a recipe job. A schedule runs the job every Monday and writes the output to S3 without a manual run. Learn more: What is AWS Glue DataBrew? (opens in a new tab)
D. Wrong: Quick Sight, formerly QuickSight, is a visualization service, and fixing data inside a dashboard is manual and hard to reuse. The option does not write cleaned data to S3 on a schedule.
Explanation
DataBrew lets analysts without Python build cleaning steps as a recipe, then run that recipe on a schedule and write the output to S3. The steps stay visible and reusable across the weekly refresh. Engineer-run notebooks, on-query views and dashboard fixes each require manual work or do not produce a cleaned output.
Want more on Data preparation with DataBrew and SageMaker? The 5 full-length practice sets have 9 more Data preparation with DataBrew and SageMaker questions, 325 in all, none of them in the free exam. One payment of $14.99, no subscription. See the practice sets →
Study Data preparation with DataBrew and SageMaker