25 Data Engineer Preguntas de entrevista

Walk me through how you'd design a data pipeline from ingestion to analytics-ready tables.

How do you choose between a data warehouse, data lake, and data lakehouse?

Describe your approach to data quality — how do you detect and handle bad data?

What's the difference between a star schema and a snowflake schema? When do you use each?

How do you handle late-arriving data in a streaming pipeline?

Describe a data pipeline you built that turned out to be harder than expected.

How do you approach incremental vs. full-load strategies for data ingestion?

What's your experience with real-time streaming — Kafka, Flink, or Spark Streaming?

How do you manage schema evolution without breaking downstream consumers?

What's your strategy for partitioning large datasets for query performance?

How would you build a pipeline that processes 1TB of data per day cost-effectively?

Describe how you test data pipelines — what does your testing strategy look like?

How do you implement data lineage and impact analysis in a complex data platform?

What's your approach to slowly changing dimensions in a data warehouse?

How do you handle PII and compliance requirements like GDPR in your pipelines?

Describe your experience with dbt or similar data transformation tools.

How would you optimize a Spark job that's running out of memory?

What's your strategy for monitoring data pipeline SLAs?

How do you approach documentation for a data platform used by non-engineers?

Describe a time you had to re-architect a pipeline because requirements changed.

How do you decide when to use SQL vs. Python for data transformations?

What's your experience with orchestration tools — Airflow, Prefect, Dagster?

How would you implement a cost-per-query tracking system for a data platform?

What's your approach to data access control and row-level security?

How do you keep data engineers aligned with data scientists and analysts?