The recruiter call set the tone. It felt transactional — boxes being ticked rather than a real conversation about the role or the team. I came away unsure whether there was much actual interest in me as a candidate, which in hindsight was probably the most honest signal in the whole process.
The task is advertised as 3–5 hours. The bar they actually review against is 3–4x that, minimum. You are not being assessed on a 3–5 hour deliverable. You are being assessed on what a senior engineer can produce in several days of evenings, against a stated budget that exists to make the ask sound reasonable.
I know this because I submitted a README with a "Future Improvements" section explicitly listing the trade-offs I'd made for scope and the things I'd do with more time — broader test coverage, auth/session consolidation, networking separation. The rejection feedback came back citing several of those exact items as reasons for the no. Other points were stylistic preferences dressed up as defects: thin service wrappers that were deliberate seams, a one-line bundle reference, view decomposition taste. None of it was discussed with me. There was no follow-up conversation, no chance to walk through trade-offs — which is the entire point of a take-home in a senior interview.
If the goal is to understand how a candidate thinks, you talk to them about the code. If the goal is to filter on whose evenings are most disposable, you do this.
The written feedback was at least detailed, which is the only reason this isn't a zero-effort rejection like the other ones described on this page. But the process itself is disrespectful of senior candidates' time, the stated time budget is not honest, and the front end of the funnel doesn't feel like the team is genuinely engaged with who they're talking to.
Advice to other candidates: If you do the task, cap yourself at the stated 5 hours, submit what you have, and move on. Don't sink your weekend into it expecting a real conversation about your code — there isn't one. Treat it as a lottery ticket, not a serious technical assessment.
Advice to management: Either raise the stated time budget to match the bar you actually apply, or assess submissions against the scope you set. Talk to candidates about their code before rejecting them on it. And train your recruiters to sound like they want to be on the call.
