Reviewed by Aditya Kumar · Last reviewed 2026-03-24
**Architecture**: Medallion or layered: Raw (Cloud Storage) → Curved (Dataflow/Dataproc) → Curated (BigQuery). **Ingest**: Pub/Sub for streaming; Cloud Storage for batch (Scheduled transfer, gsutil). **Processing**: Dataflow (Apache Beam) for streaming and batch—unified API....
This hard-level Cloud/Tools question appears frequently in data engineering interviews at companies like Tech Mahindra. While less common, it tests deeper understanding that distinguishes strong candidates. Mastering the underlying concepts (airflow, bigquery, partition) will help you answer variations of this question confidently.
This is a senior-level question that tests architectural thinking. Lead with the high-level design, then drill into specifics. Discuss trade-offs explicitly - there is rarely one correct answer. Show awareness of scale, fault tolerance, and operational complexity.
Architecture: Medallion or layered: Raw (Cloud Storage) → Curved (Dataflow/Dataproc) → Curated (BigQuery). Ingest: Pub/Sub for streaming; Cloud Storage for batch (Scheduled transfer, gsutil). Processing: Dataflow (Apache Beam) for streaming and batch—unified API. Dataproc for Spark when you need MLlib or custom libs. Orchestration: Cloud Composer (managed Airflow). Storage/Analytics: BigQuery for SQL; GCS for object storage. Why this stack: Serverless where possible; Dataflow auto-scales; BigQuery separates compute and storage. Scalability: Dataflow scales workers; BigQuery scales slots. Partition and cluster BigQuery tables by date and key columns. Cost: Dataflow and BigQuery are pay-per-use; over-provisioning Dataproc clusters is costly—use autoscaling and preemptible. Best practice: Idempotent pipelines; schema registry for streaming; RBAC via IAM and BigQuery column-level security.
Pro-Move: 'We use Dataflow templates for batch and streaming with the same Beam pipeline—write once, run both modes; saves maintenance.' Red Flag: Mixing Dataflow and Dataproc without clear criteria—leads to inconsistent patterns.
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.