interviewing at Mercury in about 3 weeks for a data engineering role. the recruiter gave me a general overview but not a lot of detail on the technical format. would love to hear from anyone who's done the data or infra loop in the last 6 months.
specifically curious about: SQL depth (window functions, performance, modeling?) any system design or pipeline design questions? do they ask about financial data or payments specifically, or is it more general? how much is behavioral vs. technical split
drop whatever you've got, even partial. I'll share back after my loop.
4 replies
ds_dmitri
not data eng but did a DS loop there ~5 months ago. SQL was serious, definitely window functions and they wanted me to write and explain a recursive CTE. financial context did come up, they asked me to think about fraud signal design which felt like it could apply to DE pipelines too.
de_derek
fraud signal design, interesting. makes sense given what they do. I'll brush up on that angle. thanks.
infra_ines
I didn't make it past the HM screen there so take this with salt, but the HM mentioned they care a lot about data reliability and observability, less about pure throughput. their stack seemed pretty standard (Airflow, dbt) but they were asking about how you handle bad data upstream, not just pipeline speed.
corp_refugee
Mercury is genuinely small so be ready for 'you'll own this entire thing' conversations. if you're coming from a big-co where you handed off to a data platform team, think through how you'd handle things end-to-end. that seemed to be the underlying theme of everything they asked.