Ctrl K
Data Engineer · Mid-level · System design

Design an analytical data warehouse for a mid-sized online store: the data comes from the product Postgres database, a CRM and ad platforms, and the consumers are BI dashboards and analysts. Walk me through how you'd set it all up.

~10 min by voice

How to approach it

The interviewer wants to see whether you can design a complete analytics pipeline end to end: from clarifying requirements and extracting data to warehouse layers, mart modeling, orchestration, data quality and access. It's not enough to name tools; you need to justify the choices and show that you're thinking about running the system in production.

Have you seen this question in a real interview?
Answers0
Avg. score—
Strong—
Practiced0
Topics
Similar questionsDesign a system of daily batch pipelines for several hundred tables with dependencies between them. How would you handle reruns, backfills and late-arriving data? SeniorYou need to deliver changes from a multi-terabyte production PostgreSQL database to an analytics warehouse with no more than five minutes of latency. How would you design this pipeline? SeniorDesign an analytical data warehouse for a marketplace: orders, products, sellers, buyers, delivery. How would you model the data, and why that way? SeniorA daily pipeline loads yesterday's data into the warehouse. Sometimes it fails halfway through, and after a restart duplicates show up in the table. How would you redesign it? Mid-levelWhat's the difference between ETL and ELT? Junior

Answer by voice

The interviewer will ask this question and listen. Think out loud and ask for a hint if you need one. When you're done, pause or tap “That's my answer”. In about a minute, your score and the points you covered show up here.

Sign in to answer

Best with headphones · ~10 min