I applied through a recruiter.

A standard recruiter chat was followed by a 60-minute technical screen that felt like a total speed test: 25 minutes for SQL (window functions, self-joins) and then 25 minutes for Python (dictionaries, sets, and list manipulation). It's less about complex logic and more about how fast you can write clean code without an IDE to help you.

The onsite was 4 rounds: three full-stack sessions and one ownership behavioral round.

• Full-stack session 1-3: Each full-stack round is a 60-minute case study where you take a product scenario, like Instagram Reels or Marketplace, and drive it end-to-end. You define the KPIs (product sense), build out the schema (data modeling), and then write the actual SQL to fetch those metrics. The modeling part is where they really grill you the most and ask follow-up questions; be ready to explain why you chose specific dimension types. The transition from a product goal to a technical schema is the hardest part.
• Ownership behavioral round: A discussion focusing on ownership.

I did a few mocks on Prepfully specifically for these full-stack rounds, and it helped me get the timing down. I'd say know your Kimball fundamentals (SCDs, bridge tables) and be mindful of time.

Design a data model for a new housing marketplace feature. Design the schema and write the SQL to calculate the conversion rate from view to lead for each location.

1 answer.