Netflix Senior Software Engineer Interview Experience — Onsite Loop in the Ads Org: Topo Sort, System Design, and Two BQs

I passed this earlier this year, for the Ads department, which has been pretty popular lately. Since I'm planning to accept the offer, I'm not going to post the specific questions — search for the last three months of posts here and you'll find they're almost all the same type. I'll share some experience and feelings that I hope will be useful. Process-wise: they got back to me 1 day after the phone screen, and 2 days after the onsite.

Phone screen: topological sort. Got the result the next day (usually within two days).

Onsite. The onsite questions were almost identical to what others have posted before. The most common coding questions posted recently are topological sort, system design (freq cap), and data model, plus two rounds of BQ, spread across two days.

Coding — topological sort. Not much to say about the coding topo question. It's not the exact same problem, but it's essentially the same thing. One extra thing to keep in mind: know the runtime cold, and prepare for variations — like sorting multiple results, or printing all results along with their runtimes.

System design (tests depth). They dig pretty deep here — or if the interviewer doesn't dig deep, dig deep yourself. Things like: how is Redis implemented, what modes does it have, how do you handle traffic, how do you make it atomic (and why you'd need to). Cover both the online and offline path at a high level (read/write). What MQ do you use on the offline path, how do you design the key, and if there's a special requirement like strictly enforcing 1 ad per cap, how would you implement that.

Data model (tests breadth). Go off what you've actually worked on, or what you learned from GPT. Ideally cover all the major parts of an ads system — ad campaign, business account, creative, user targeting — write down everything you can think of, and talk through how they connect. (I also talked through the overall ads platform flow — intake flow, serving flow, impression tracking flow — and then went deeper into the intake part. That's probably not strictly necessary, it just shows you understand the whole system.)

BQ x2 (tests whether your work experience really matches the JD — can you hit the ground running). Both rounds are with managers, so there's not much to say about culture fit. I don't think people get cut in this round on the common questions. The manager round is basically with a manager who's interested in you — if you know which team or which role you're interviewing for, read the job description carefully, make sure you've actually done 80%+ of it, and lean your experience talk in that direction. The manager just wants someone who's already done this stuff and can start working right away.

They told me the result about two days later. Since I got along really well with the manager, it went smoothly — what they need done is exactly what I've been doing, so during the interview the manager basically hinted that as long as the other rounds passed, I'd be brought straight in.

Some experience I think is worth sharing: the Ads department has been doing centralized hiring recently, with its own question bank and its own criteria. I'd say it's basically open-book and not that hard in itself — for whatever parts you haven't done, just prepare more with GPT. What they're really hiring for is people with relevant experience who can start working right away, so it really helps to make sure the role and team you're interviewing for matches your own experience by 80%+. The manager round is probably the most important — if that conversation goes well, as long as the position is still open, you should be fine.