NewML System Design & SQL tracks just landedSee what's new
avatar
dingwei91Jun 29, 2026

Ramp onsite - Rejected and still don't know what signal they were looking for

Interview Experiences
ramp
5210858

Replies (7)

avatar
priya.nairJun 29, 2026

The silence during coding is so on brand for Ramp. My interviewer literally just stared at his screen the whole time. No indication of good or bad.

26
avatar
User-j2kzmJun 29, 2026

How long was your phone screen to onsite turnaround? Wondering if speed of scheduling says anything about how they view you.

20
avatar
dingwei91Jun 29, 2026

Next business day call from recruiter, onsite scheduled about a week out. So pretty fast. Didn't seem to matter in the end.

10
avatar
rohitsharma7Jun 29, 2026

4 YOE and went through all 4 rounds, that's already a solid filter. Most folks I know didn't even get past the phone screen. Still sucks to not get an offer tho, sorry man.

19
avatar
User-ric39Jun 29, 2026

For the flight problem can I use TS instead of python? What is the problem with the file import and tests, I want to prepare for this problem friction in advance. Thank you very much!

7
avatar
User-9rj1lJun 30, 2026

Did you get any feedback at all or just the standard "we've decided to move forward with other candidates" email?

4
avatar
User-jar39Jun 29, 2026

Same experience with the architecture round. Recruiter said project deep dive only and then surprise system design showed up. Why can't they just tell you the format upfront?

3
avatar
dingwei91Jun 29, 2026

Yeah the recruiter miscommunication is what got me most. If I'd known there was going to be a system design I would have practiced more of those. My project walkthrough was solid, and then the SD just threw me off rhythm.

16
avatar
User-ric39Jun 29, 2026

Reply to @dingwei91Seems like Ramp loves to throw curve balls, impossible to prepare. I am worried now about the non-ai coding round friction around the file import setup, could you please guide me through that?

1
avatar
User-ave24Jun 29, 2026

The AI coding round is a black box. Nobody I've talked to could figure out what they were measuring. Like are they testing your AI usage? your problem decomposition? the matching logic? Genuinely no idea.

1
No more replies