Real questions from top companies
What role does executor memory and CPU configuration play in maximizing parallelism?
What role would Kafka or similar event-driven platforms play in your architecture?
What strategies would you use to optimize Spark jobs for both performance and cost on AWS?
What strategies would you use to reduce latency in a streaming data pipeline?
What techniques ensure deduplication in large datasets?
What trade-offs would you consider when choosing between batch processing and real-time streaming?
What's the difference between narrow and wide transformations?
When submitting Spark jobs, how does the process work in the backend? Explain.
When would you choose a broadcast join over a shuffle join? Any memory risks?
Which Spark property controls the number of shuffle partitions?
Which Spark version are you using in your project, and why did you choose it?
Why I chose specific technologies (e.g., Spark over traditional ETL tools)
Why does Hive use Derby by default, and what alternatives are used in production?
Worked with UDFs - share examples
Write PySpark code to extract data from a CSV and create a table.
Write PySpark code to filter and count records.
Write PySpark code to filter records based on specific conditions and add a calculated column.
Write PySpark code to save a DataFrame in Parquet format to an S3 bucket.
Write a PySpark code snippet to filter rows with a specific condition.
Write a PySpark job that calculates the number of unique users who logged in per day, but exclude any logins from inactive users listed in a separate file.
Type or paste your answer to any of these questions and our AI Coach scores it, highlights gaps, and rewrites it at FAANG quality. Free to try.