Mercury · Primly Community

Mercury onsite / final round, how it really goes

mobile_mara · 5 replies

Finished my Mercury final round loop about three months ago. Senior SWE role, remote. Going to write this up properly because I had to piece together info from four different forum posts and none of them were complete.

The full onsite was five rounds over two days (remote, but blocked scheduling). Here's the lineup:

Coding (45 min): Live coding, one problem. I described mine in another thread. Medium-hard level. Interviewer wanted clean, readable code and asked me to walk through edge cases out loud.

System Design (60 min): Fintech-flavored. Design a payment processing system was the gist. They push on failure modes, idempotency, eventual consistency. See other posts on this thread for more detail.

Behavioral / Values (45 min): Two interviewers, co-interview style. One asked, one took notes. Questions tied to their stated values. They want stories with specific outcomes.

"Bar Raiser" round or cross-functional (45 min): At Mercury this felt like a senior engineer or EM outside the immediate team asking broader questions. Mix of technical judgment ("how would you approach a codebase that's accumulated a lot of tech debt") and behavioral.

Hiring manager (30 min): More conversational. Partially about the role, partially about career trajectory. They're figuring out if you'll stick around and grow.

Debriefs: I heard back with a decision in 4 business days. They give written feedback if you don't get the offer, which is rare and actually appreciated.

One thing worth knowing: they care about writing. There's a written portion in some track variants where you respond to a hypothetical scenario in text form before the loop. Not everyone gets this, might depend on role. If you do, don't rush it. They read it carefully.

Offer came in above my expectation given the stage and size. Base was competitive with Series D/E market. Equity was straightforward, no weird cliff games.

5 replies

content_cole

The written component is real. I got it and underestimated it. Treat it like a design doc, not a slack message. They're testing communication quality, not just content.

frontend_fran

Five rounds across two days sounds tiring. Did you feel like you had enough breaks or was it back-to-back?

staff_steph

They did block 30-min gaps between each. Appreciated it. Still exhausting, but not inhumane.

alex_design

Did anyone ask a take-home coding challenge instead of the OA, or is OA universal?

quietquit_quincy

I got the OA. Heard from one other person who got a take-home instead. Might be role or team dependent.