AWS Certified Data Engineer – Associate (DEA-C01) tests whether you can build and run data pipelines on AWS: ingesting batch and streaming data, transforming it with Glue, EMR or Lambda, storing it in S3, Redshift or DynamoDB, orchestrating with Step Functions or Airflow, and keeping it secure and governed. The current exam guide also covers vector indexes (HNSW, IVF), Apache Iceberg, using LLMs for data processing and Amazon SageMaker Unified Studio.
This guide follows AWS's official exam guide: all four domains with weights, the service choices that decide most questions, a 7-week plan and 15 original scenario questions.
Quick facts (AWS, November 2026)
| Item | Detail |
|---|---|
| Exam code | DEA-C01 |
| Time | 130 minutes |
| Questions | 65 (50 scored, 15 unscored), multiple choice and multiple response |
| Passing score | 720 on a 100–1,000 scale |
| Fee | USD 150 (plus applicable taxes and exchange charges) |
| Languages | English, Japanese, Korean, Simplified Chinese |
| Valid for | 3 years |
| Recommended experience | 2–3 years in data engineering or data architecture, plus 1–2 years hands-on with AWS |
Read the official DEA-C01 page and exam guide before booking. AWS publishes guide revisions at least one month before they appear on the exam, so check the Revisions section too.
Who should take DEA-C01?
- ETL, SQL and BI developers moving to AWS data platforms.
- Data engineers working with Glue, EMR, Redshift, Athena or Kinesis.
- Python developers and analysts who build pipelines for analytics or AI.
AWS lists ML model training, language-specific syntax and drawing business conclusions as out of scope. Freshers should first learn SQL, Python and AWS basics. The Cloud Practitioner is a gentle start, and the data engineer career path shows the full route.
Domains and weights
| Domain | Weight |
|---|---|
| 1. Data Ingestion and Transformation | 34% |
| 2. Data Store Management | 26% |
| 3. Data Operations and Support | 22% |
| 4. Data Security and Governance | 18% |
Domain 1: Ingestion and transformation (34%)
- Ingestion: streaming from Kinesis, Amazon MSK, DynamoDB Streams and DMS; batch from S3, Glue, EMR, Redshift and AppFlow; schedulers with EventBridge or Airflow; S3 event triggers; Lambda from Kinesis; throttling and rate limits; fan-in and fan-out; replayability; stateful vs stateless.
- Transformation: Glue, EMR, Lambda and Redshift; JDBC/ODBC sources; CSV to Parquet; cost-aware processing; troubleshooting failures; data APIs; LLMs for data processing.
- Orchestration: Step Functions, Amazon MWAA (Airflow), Glue workflows, EventBridge and Lambda; resilient, serverless pipelines; SNS and SQS alerts.
- Programming: Python, SQL and Scala; Lambda concurrency; Git, testing and logging; CloudFormation, CDK and AWS SAM; CI/CD for pipelines; distributed computing basics.
# AWS Glue (PySpark): read raw CSV from the catalog, write partitioned Parquet
from awsglue.context import GlueContext
from pyspark.context import SparkContext
glue = GlueContext(SparkContext.getOrCreate())
orders = glue.create_dynamic_frame.from_catalog(database="raw", table_name="orders_csv")
glue.write_dynamic_frame.from_options(
frame=orders, connection_type="s3", format="parquet",
connection_options={"path": "s3://datalake-curated/orders/", "partitionKeys": ["order_date"]},
)Domain 2: Data store management (26%)
- Choosing stores: Redshift, EMR, Lake Formation, RDS, DynamoDB, Kinesis Data Streams and MemoryDB; HNSW indexes in Aurora PostgreSQL; Redshift federated queries, materialized views and Spectrum; locks; Apache Iceberg; vector index types (HNSW, IVF).
- Catalogues: Glue Data Catalog and Hive metastore; crawlers; partition sync; business catalogues in Amazon SageMaker Catalog.
- Lifecycle: COPY and UNLOAD between S3 and Redshift; S3 Lifecycle tiering and expiry; S3 versioning and DynamoDB TTL; deleting data for legal needs.
- Modelling: schemas for Redshift, DynamoDB and Lake Formation; schema evolution; AWS SCT and DMS Schema Conversion; lineage; partitioning and compression; vectorisation with Bedrock Knowledge Bases.
| Need | Usual choice |
|---|---|
| Ad-hoc SQL on files in S3, pay per query | Athena |
| Data warehouse with heavy joins and BI dashboards | Redshift |
| Query S3 data from Redshift without loading it | Redshift Spectrum |
| Single-digit-millisecond key-value lookups at scale | DynamoDB |
| Real-time event stream with multiple consumers | Kinesis Data Streams or Amazon MSK |
| Serverless Spark ETL with a catalogue | AWS Glue |
| Large custom Spark or Hadoop clusters | Amazon EMR |
| Fine-grained table and column permissions on the lake | Lake Formation |
Domain 3: Data operations and support (22%)
- Automation: MWAA and Step Functions; SDK calls; EMR, Redshift and Glue features; Glue DataBrew and SageMaker Unified Studio for preparation; Athena queries; Lambda and EventBridge.
- Analysis: QuickSight and DataBrew; SQL views in Redshift and Athena; Athena notebooks with Spark; provisioned vs serverless trade-offs; aggregation, rolling averages, grouping and pivoting.
- Monitoring: audit logs, CloudTrail, CloudWatch Logs and Logs Insights, OpenSearch; troubleshooting Glue and EMR jobs.
- Data quality: checks during processing, DataBrew rules, consistency, sampling and handling data skew.
-- Athena: daily revenue with a 7-day rolling average from partitioned Parquet
SELECT order_date,
SUM(amount) AS revenue,
AVG(SUM(amount)) OVER (ORDER BY order_date ROWS BETWEEN 6 PRECEDING AND CURRENT ROW) AS revenue_7d
FROM curated.orders
WHERE order_date >= date '2026-10-01'
GROUP BY order_date
ORDER BY order_date;Domain 4: Data security and governance (18%)
- Authentication: security groups; IAM roles for Lambda, API Gateway and CloudFormation; Secrets Manager rotation; S3 Access Points and PrivateLink; SageMaker Unified Studio domains and projects.
- Authorisation: custom least-privilege policies; Redshift users, groups and roles; Lake Formation permissions across Redshift, EMR, Athena and S3; role-, tag- and attribute-based access.
- Encryption and masking: KMS, cross-account encryption, encryption in transit, masking and anonymisation.
- Audit logs: CloudTrail, CloudTrail Lake, CloudWatch Logs, Athena and OpenSearch analysis.
- Privacy and governance: Redshift data sharing; PII detection with Macie; stopping backups or replication to disallowed Regions; AWS Config; data sovereignty.
For Indian companies, data-sovereignty questions usually mean keeping data in Mumbai (ap-south-1) or Hyderabad (ap-south-2). See also the DPDP Act for AI applications.
7-week study plan
| Week | Focus | Lab |
|---|---|---|
| 1 | S3 data lake layout, Glue Data Catalog, crawlers, Athena | Crawl raw CSV, query it with Athena, convert to Parquet |
| 2 | Glue ETL jobs, DataBrew, Iceberg tables | Partitioned Iceberg table with an upsert |
| 3 | Kinesis Data Streams, Firehose, MSK, Lambda consumers | Stream clicks into S3 with Firehose; Lambda enrichment |
| 4 | Redshift: COPY/UNLOAD, distribution and sort keys, Spectrum, data sharing | Load curated data and build a materialized view |
| 5 | Step Functions, MWAA, EventBridge, SNS alerts; DynamoDB design | Orchestrated daily pipeline with a failure alert |
| 6 | Lake Formation, KMS, Macie, CloudTrail Lake, Config | Column-level permissions that hide PII from analysts |
| 7 | Vector indexes, Bedrock Knowledge Bases, official practice questions | Review wrong answers; book the exam |
15 scenario questions with answers
These are original scenarios written for this guide, not exam questions.
1. Analysts need SQL on raw files in S3 with no servers and pay-per-query pricing. What do you use?
Amazon Athena with the Glue Data Catalog.
2. Athena queries scan too much data on large CSV files. What helps most?
Convert to a columnar format (Parquet), compress, and partition by common filters such as date.
3. New files arrive in S3 at unpredictable times and must be processed immediately. How?
S3 Event Notifications (or EventBridge) triggering Lambda or a Glue job.
4. A Lambda consumer can't keep up with a Kinesis stream. What do you check?
Shard count, batch size and parallelisation factor, and Lambda concurrency limits.
5. A pipeline has ten steps with retries, branches and a human-readable workflow. What orchestrates it?
AWS Step Functions (or Amazon MWAA if the team already uses Airflow DAGs).
6. You need ongoing replication from an on-premises Oracle database into S3. What service?
AWS DMS with change data capture.
7. Redshift must query historical data in S3 without loading it. What feature?
Redshift Spectrum.
8. Glue crawlers keep creating many small tables instead of one partitioned table. What do you fix?
The S3 folder layout and crawler grouping settings so the files share a consistent schema and partition structure.
9. A lake table needs row-level updates, deletes and time travel. Which table format?
Apache Iceberg.
10. Analysts may see order data but not the customer phone column. How?
Lake Formation column-level permissions (or data masking at query time).
11. You must find which S3 buckets contain PII. What service?
Amazon Macie.
12. Raw logs must move to a cheaper tier after 30 days and be deleted after a year. How?
S3 Lifecycle rules for transition and expiration.
13. Encrypted data in one account must be readable by a job in another account. What do you configure?
The KMS key policy to allow the other account's role, plus IAM permissions and the bucket policy.
14. A pipeline must reject batches with empty customer IDs before loading. What do you add?
Data quality rules (Glue Data Quality or DataBrew) that fail or quarantine the batch.
15. A support chatbot needs semantic search over product manuals stored in S3. What fits?
A Bedrock Knowledge Base (or embeddings in a vector store such as Aurora PostgreSQL with an HNSW index).
After DEA-C01
- Practise with data engineering interview questions, AWS EMR interview questions and Databricks interview questions.
- Compare the Google path in our Professional Data Engineer guide.
- See live data engineering jobs.
Cloud Soft Solutions in Ameerpet, Hyderabad trains data engineering with SQL, Python, Spark and AWS data services. See the data engineering and AI/ML course and all courses.
📞 Call +91 96660 19191 · WhatsApp · Flat No. 513, 5th Floor, Nilagiri Block, Aditya Enclave, Ameerpet, Hyderabad
Frequently Asked Questions
How long is the AWS DEA-C01 exam?
130 minutes, with 65 questions: 50 scored and 15 unscored.
What is the passing score for DEA-C01?
720 on a scaled score of 100 to 1,000.
How much does the AWS Data Engineer Associate exam cost?
USD 150, plus applicable taxes and exchange charges.
Which DEA-C01 domain has the most weight?
Data Ingestion and Transformation, at 34% of scored content.
What experience does AWS recommend for DEA-C01?
2–3 years in data engineering or data architecture and 1–2 years of hands-on AWS experience.
Does DEA-C01 cover vectors and AI?
Yes. The guide includes vector index types such as HNSW and IVF, Bedrock Knowledge Bases and using LLMs for data processing.
