have a computational role loop coming up and i want to calibrate. everything i'm finding is from 2022-2023 and i'm not sure how much has changed.
if you've been through the DE Shaw process in the last 12 months, would really appreciate: what track (computational vs quant), how many rounds, what the coding difficulty felt like, whether systems design came up, and roughly how long the full thing took.
also any sense of whether the behavioral portion is just a vibe check or something you actually need to prep for specifically.
drop whatever you have, even partial loops are useful.
4 replies
numbers_only
went through it 4 months ago, computational track. 5 rounds: HackerRank screen, 2x live coding, 1x systems design, 1x behavioral. coding was solidly LeetCode hard territory, systems design focused on data pipelines and consistency tradeoffs. behavioral was 2 interviewers and legitimately substantive, they pushed back on my answers. total time: 8 weeks.
ml_mike
quant track, same timeframe. different shape: 1x coding screen, 1x probability/stats deep dive, 1x ML system design, 1x research discussion about past projects. no LeetCode grind needed but you need to be very clean on probability and estimation. the research discussion was the hardest part honestly.
de_derek
this is really helpful. i'm computational so the LeetCode hard framing is useful calibration. 8 weeks is also what i was expecting. thanks both.
staff_steph
for the behavioral, prep beyond just having stories ready. they want to understand WHY you made the decisions you did, not just what happened. think less STAR and more 'decision journal entry.'