got a recruiter screen scheduled for a data engineering / platform role at Cohere and would love to hear from anyone who's been through a loop there recently, ideally in the last 6 months.
specifically curious about: how many rounds, how technical is the system design portion, any SQL or data modeling questions or is it more distributed systems focused, and what does the behavioral portion actually look like.
the JD was a bit vague on process. any data points appreciated, i'll share my own when i come back through.
3 replies
ml_mike
MLE loop so slightly different but the system design round for me was heavily distributed inference / serving. i'd assume data eng would be similar, they care about how data moves at scale and latency. probably less SQL trivia, more pipeline design.
infra_ines
a friend did a platform eng role there last fall. she said the system design round involved a scenario around data ingestion for fine-tuning pipelines. not a general 'design Twitter' thing. more applied to what they actually build. i'd think about how training data gets cleaned, versioned, and served.
ds_dmitri
the behavioral section in AI companies tends to focus on 'how do you decide when a model is good enough to ship.' if you've had that conversation at a previous job, have a crisp answer ready.