Redesign a clickstream pipeline so the next latency spike, schema drift, credential rotation, downst
A medium Pipeline Design mock interview question on DataDriven. Practice with AI-powered feedback, real code execution, and a hire/no-hire decision.
- Domain
- Pipeline Design
- Difficulty
- medium
Interview Prompt
Redesign a clickstream pipeline so the next latency spike, schema drift, credential rotation, downstream outage, and producer-side burst do not cause an outage. Compose every advanced-tier pattern: failure-surface classification on each node; retry budget plus circuit breaker around every external call; DLQ with envelope; DLQ-as-quality-signal monitoring (rate by exception type with z-score alerts); DLQ replay tooling with a second-failure side channel; bounded queues with backpressure; load-shedding on the source side. Route the dashboard through the resilient path.
How This Interview Works
- Read the vague prompt (just like a real interview)
- Ask clarifying questions to the AI interviewer
- Write your pipeline design solution with real code execution
- Get instant feedback and a hire/no-hire decision