Reviewed by Aditya Kumar · Last reviewed 2026-03-24
**Architectural decision framework**: Choose by data volume, latency, complexity, and operational model. **Lambda**: Event-driven, <15 min, <10 GB memory, stateless. Use for: S3/Kinesis/EventBridge triggers, lightweight transforms (JSON reshape, validation), API backends,...
This medium-level Cloud/Tools question appears frequently in data engineering interviews at companies like Bitwise. While less common, it tests deeper understanding that distinguishes strong candidates. Mastering the underlying concepts (etl, join, spark) will help you answer variations of this question confidently.
Break this problem into components. Identify the core trade-offs involved, then walk the interviewer through your reasoning step by step. Demonstrate awareness of edge cases and production considerations - this is what separates good answers from great ones.
Architectural decision framework: Choose by data volume, latency, complexity, and operational model. Lambda: Event-driven, <15 min, <10 GB memory, stateless. Use for: S3/Kinesis/EventBridge triggers, lightweight transforms (JSON reshape, validation), API backends, orchestration (Step Functions). Suits sub-GB data and sub-second to low-second latency. Cost: Pay per invocation + GB-second; at 1M invocations/month with 512 MB × 3 s, ~$20. Glue: Managed Spark, batch ETL, large datasets (TB scale), Data Catalog integration, crawlers. Use for: nightly batch jobs, complex joins, Parquet conversions, schema evolution. Suits hours of runtime. Cost: DPU-minute; 10 DPU minimum = ~$0.44/run even for 1-min job. Trade-off: Lambda is cheaper for small, frequent work; Glue for large, infrequent. At 100 MB/event × 1000 events/day, Lambda may thrash; Glue batch is better. In production: Lambda for real-time ingestion and triggers; Glue for daily aggregations and heavy ETL.
Pro-Move: 'We use Lambda for S3→validation→forward, and Glue for nightly fact-table builds—the boundary is 1GB and 5-minute complexity threshold.' Red Flag: Using Glue for sub-minute triggers or Lambda for multi-GB processing—wrong tool for the job.
Some links below are affiliate links. If you buy through them we may earn a small commission at no extra cost to you — it helps keep DataEngPrep free.
According to DataEngPrep.tech, this is one of the most frequently asked Cloud/Tools interview questions, reported at 1 company. DataEngPrep.tech maintains an editor-reviewed database of 1,863 data engineering interview questions across 7 categories.