Real Interview Experiences
Learn what to expect directly from candidates and interviewers who've been through it.
- 1,326 experiences
- 314 companies
- 37 roles
- +207 added this month
OpenAIResearch Engineer“The coding plus ML stats round was the hardest by far. They basically asked me to implement an all_gather on noisy nodes, derive how many rounds you’d need for a target error, then figure out a better algorithm using the fact you’re transmitting floats.”
Harvey AIPrincipal MLOps Engineer · L8+“I ended up pitching this quantum ML architecture where we were training models on quantum hardware to do something as stupidly simple as predict whether a number was even or odd, and Harvey’s panel got weirdly fascinated by the remote coding environment I built for physicists.”
NvidiaSenior MLOps Architect · L5“They showed me this intentionally vague architecture with Kafka, Jupyter, and an experiment dashboard, and the whole prompt was basically just, "it’s slow." Then they pivoted to a custom system training LLMs across 500 nodes and kept drilling into failure modes and latency.”
Sierra AIAgent Engineer“I told Sierra I’d only worked in Python, they said that was fine, and then my first onsite was a TypeScript React debugging round where the interviewer showed up 10 or 15 minutes late and was super rude and cold.”
RedditStaff Machine Learning Engineer · L6“I interviewed at maybe 15 or 20 places in six months, and for every onsite I got an offer, but Reddit still threw me for a loop because the ML system design leaned way more infra than modeling.”
IntercomMachine Learning Engineer · L5“Well structured. Before every round they gave us prep material and a mock round with a recruiter. Interviewers are knowledgeable and they are clear in what they want.”
LinkedInMachine Learning Engineer · L5“The coding round was the most tricky one, you don't know what problem you will get for the leetcode round. ML fundamentals, ML design is easy if you prepare for regular ML system designs. AI coding, do not use the cutting edge models, it takes forever to run. Make sure to downgrade your model, so it answers fast. Behavioral is pretty standard.”
DoorDashMachine Learning Engineer · L4“The overall process is okay. Although there were multiple interview rescheduling for the onsite. I feel I did not do well for the ML System Design round. The question was asking how to design a chat bot for merchants, which is something I did not prep for. They would ask you to draw out design, how to make modeling choices, and etc. Pretty detailed oriented so be prepared to speak on many of that fronts. ”
GoogleMachine Learning Engineer · L4“2 coding rounds, 1 googliness round, 1 AI/ML depth. THe AI/ML was mostly going over resume, recent projects and some ml system design”
AmazonApplied Scientist · L4“The questions are not difficult but do require some actual experiences. Not much preparation can help.”
Preview full access of other recent interview experiences
AnthropicSoftware Engineer, Safeguards · L5“The weirdest Anthropic round was the company values interview. It was almost like a therapy session, and honestly if you went to a therapist at some point, you will pass that round much more easily.”
NetflixEngineering Manager, Tooling“What was very unusual is they didn’t give me any tooling to draw the system design, so I just sketched it on a piece of paper and talked them through it, then we got into this oddly deep debate about whether hover-over history should count as a recommendation signal.”