I agree with a lot of this, Marcus. Interviewing practices have changed.
I think cultural fit is important for easy team/company alignment but, recently, I read someone suggesting otherwise, that differences or misfits are good for innovations and checking norms. Curious where you stand on that?
That's very good point, and I do agree. We don't need the "same mindset" all over the crew, but sharing the core values so we ensure we are all aligned so no one gets frustrated.
A common scenario I saw: MVP vs All done.
If the team/company values are around making MVP and, from there, increase the value to "All done", but he way of working for a new employee is quite the opposite, you have a hiring problem:
That's a good example. Personally, I've seen it when people come from big orgs with heavy processes to a smaller companies with less formal structure. If they don't have much agency it can be very disorientating for them.
I think "mindset" is the right differentiator here.
Really good article! I completely agree with your take on how interview practices are changing. One trend I've noticed, both as a candidate and as an interviewer, is replacing traditional coding exercises with a deep dive into a project the candidate built with AI. Instead of focusing on writing code from scratch, the discussion revolves around system design, architecture, project structure, tech stack, AI workflows, and, most importantly, the reasoning behind the decisions that were made.
From the interviewer's perspective, these sessions reveal something else too: how people actually work with AI. Some candidates accept AI-generated changes almost blindly, leaving behind inconsistent naming, architectural mismatches, or even exposed credentials. Others use AI as a true engineering assistant: the code feels cohesive, every decision is intentional, and it's hard to tell where AI stopped and the engineer took over. To me, that's becoming a much more valuable signal than whether someone can solve a coding puzzle in 30 minutes.
I agree with a lot of this, Marcus. Interviewing practices have changed.
I think cultural fit is important for easy team/company alignment but, recently, I read someone suggesting otherwise, that differences or misfits are good for innovations and checking norms. Curious where you stand on that?
That's very good point, and I do agree. We don't need the "same mindset" all over the crew, but sharing the core values so we ensure we are all aligned so no one gets frustrated.
A common scenario I saw: MVP vs All done.
If the team/company values are around making MVP and, from there, increase the value to "All done", but he way of working for a new employee is quite the opposite, you have a hiring problem:
- the new employee will get frustrated.
- your team/company will not deliver
- at some point the new employee will leave
That's a good example. Personally, I've seen it when people come from big orgs with heavy processes to a smaller companies with less formal structure. If they don't have much agency it can be very disorientating for them.
I think "mindset" is the right differentiator here.
Really good article! I completely agree with your take on how interview practices are changing. One trend I've noticed, both as a candidate and as an interviewer, is replacing traditional coding exercises with a deep dive into a project the candidate built with AI. Instead of focusing on writing code from scratch, the discussion revolves around system design, architecture, project structure, tech stack, AI workflows, and, most importantly, the reasoning behind the decisions that were made.
From the interviewer's perspective, these sessions reveal something else too: how people actually work with AI. Some candidates accept AI-generated changes almost blindly, leaving behind inconsistent naming, architectural mismatches, or even exposed credentials. Others use AI as a true engineering assistant: the code feels cohesive, every decision is intentional, and it's hard to tell where AI stopped and the engineer took over. To me, that's becoming a much more valuable signal than whether someone can solve a coding puzzle in 30 minutes.
Wow! great insights from first hand, I’m pretty sure many of the readers will find this super useful! Thanks Stefania for sharing!