An operational SLA states 'fresh by 6am'; a quality SLA states 'correct row counts and null rates by
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
An operational SLA states 'fresh by 6am'; a quality SLA states 'correct row counts and null rates by 6am.' The two are independent: a pipeline can meet 6am with a 30 percent row-count drop, or have flawless data and miss 6am. Status pages that report a single number describe operational SLA exclusively, leaving consumers unable to distinguish 'late but correct' from 'on time but wrong.' Split the SLAs by adding two separate monitoring paths off the curated table, one whose name states the operational SLA target and the orchestrator metric it reads, and one whose name states the quality SLA target and the gates it depends on. Each path routes to its own paging destination.
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