Three environments serve three different validation goals
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
Three environments serve three different validation goals. Dev catches typos and broken refs; CI catches schema regressions and unit-test failures; staging catches integration issues and scale problems that slim CI does not. The section's data-shape choice per environment: dev gets sample data committed to the repo (tens to thousands of rows); CI gets a slim-CI subset of recent prod (modified models plus descendants); staging gets a subset of prod (last 7 days) or masked full prod with PII columns hashed. Choose the env data shape by adding three environment storage nodes downstream of a masking pipeline whose name states the data shape and PII handling for each.
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