![[Your Ryan] what are your honest thoughts on this manhwa?](https://preview.redd.it/yv1qg5s9xa2h1.jpeg?auto=webp&s=8be4e5eeae07370818cbd726e2fa6cb00c3a978c)
[Your Ryan] what are your honest thoughts on this manhwa?
is this worth waiting forrrr?
![[Your Ryan] what are your honest thoughts on this manhwa?](https://preview.redd.it/yv1qg5s9xa2h1.jpeg?auto=webp&s=8be4e5eeae07370818cbd726e2fa6cb00c3a978c)
is this worth waiting forrrr?
Managing a backend team at a \~30 person
software startup. hiring picked up again and I’ve been rethinking our screening/shortlisting process.
Over the past year we hired a couple developers who absolutely crushed interviews (coding, take-home, etc.) but struggled once they were inside our actual messy codebase.
Nothing dramatic, but things like tracing a failing service or debugging logs took way longer than expected.
Our old process was pretty standard. Resume screen, HackerRank/leetcode challenge, then face-to-face interview and system design/whiteboard with a tech team member. Probably \~5-6 hours of total interview time per candidate on our side.
For our last backend role we tried something slightly different. 1st round was still a small HackerRank style challenge just to filter obvious mismatches. Out of 40 applicants, 22 passed that step.
Then, instead of jumping straight to interviews, we ran a short debugging task where candidates had to fix a broken API service and explain what they were doing. The interesting signal was watching how they explored logs and unfamiliar code.
We used a platform called Utkrusht for this because it runs tasks in a live prod environment and then records the entire session. Candidates could use docs, ChatGPT, whatever they normally use.
What surprised me: the fastest candidate wasn’t the one we hired. The person we ended up hiring spent almost half the time just investigating root cause and narrating their thinking. Way slower, but much more like how real production debugging actually looks.
Because we could watch the session recording, our follow-up interviews dropped to just \~2 hours total instead of the usual 5–6. We still talked architecture and team fit, but a lot of the “how do you debug things” questions were already answered.
For people here, how are other engineering managers handling this lately? Are people still relying heavily on phone screens/coding tests, or moving toward more environment / debugging style candidate assessments?