has the QR interview actually changed since LLMs, or is it still the 2023 process?
3 YOE at a mid-size multistrat. most of my implementation work is LLM-assisted at this point and it's made me wonder if interviews have caught up to that or if they're still screening for the exact same stuff they were three years ago.
want to hear what people are actually running into, candidates and interviewers both. stuff I'm wondering:
- take-homes. are they dead? proctored? still going as normal?
- remote screens. anyone moved back to in-person, or gone the lockdown browser / share-your-whole-screen route?
- has anyone had a round where you're supposed to use a model and they're watching how you prompt it and check its work? or is it still just banned
- has the content shifted at all. more stats derivation and research taste, more "here's a broken model, find the bug", or is it still leetcode + brainteaser + stats like always
- interviewers: what's failing people now that wouldn't have two years ago?
- junior headcount. has the early-career funnel actually shrunk, or is it just skewing toward people who've already sat on a research seat somewhere?
if you can, drop your firm tier (T1 = JS/Citadel/HRT/Jump/Optiver/SIG type, T2 = large multistrat/quant shop, T3 = smaller/regional), region, and whether you're answering as a candidate or interviewer. one-line answers to any of these are fine, don't feel like you need to do all six.