A product manager at a retail analytics company asks for a real-time inventory dashboard
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
A product manager at a retail analytics company asks for a real-time inventory dashboard. The pipeline on the canvas already has the four roles but no slaFreshness labels, and no honest tier conversation has happened. The truth: storefront POS systems batch-export every 15 minutes, the warehouse mart refreshes hourly, and the Looker dashboard auto-refreshes every minute. Apply the five-tier vocabulary this section just taught and translate the real-time request into the concrete tier the chain can actually deliver. Tag each node with a slaFreshness value consistent with the slowest upstream hop: the POS source is < 15min (Tier 2), the ingest job and the warehouse mart cannot beat the 15-min POS rhythm but the warehouse refreshes hourly so it carries < 1h (Tier 3), and the Looker dashboard inherits the warehouse's tier (< 1h or < 2h). No node may carry a real-time or < 1min label, because the upstream cannot back it.
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