First I had a recruiter call, then hiring manager call talking about past experience, then coding interview where you tell an engineer what to do their machine. Overall silly. Not sure what else to say here.
Interview questions [1]
Question 1
You’ll be given a silly question to debug something in their codebase. But you need to tell another engineer sharing their screen what to do, so it just takes ages. On top of that the problem you’ll need to debug will something really silly like: an interface that is implemented by dozens of classes with different overrides. There is a reason why React no longer uses classes but functions and why most front end applications don’t rely OOP. It’s because it is because debugging them is tedious and absurd. You should not structure a FE application the same way you do BE service written in Java.
I applied online. I interviewed at Ashby in Mar 2026
Interview
1. 30 minute with a recruiter, was very friendly and helpful
2. 25 minutes (not 30) with EM. He seemed in a hurry and it just did not make any sense to make this interview just 25 minutes when you need to explain a lot of stories.
3. Pair coding full-stack interview asking you to change something in their repo of both BE/FE so if you are not senior React/GraphQL/TS/Nodejs don't even waste your time. They even had SQS included.
Pros:
Every stage included an email with a lot to read and the exact questions you will be asked. They really only asked these questions (with a bit more clarification questions). I found that very helpful and neurodivergent friendly.
Cons:
The interviewers seemed to be under time pressure always skipping the intro to jump into the questions. The coding challenge is way too overwhelming for a neurodivergent person jumping between so many tabs and trying to speak every possible thought instead of just doing it. I don't feel this kind of exercise reflects real life. Let's say you are solving an incident, you usually have more than 1 hour and the incident won't include full stack you are not familiar with and it won't include pair programming. To make this even worse, the interviewer cut me short after 38 minutes saying I won't have enough time to finish so he has to finish here. I do technical interviews at my current employer and I understand when someone doesn't perform on the level we want, but we usually keep going and give the candidate a few minutes to ask questions and end on good terms. You never know, maybe this person applies again when they fill a certain gap or interview you one day in the future. Keeping it professional reflects the image of the company. In this case, these interviewers only gave a negative image and that their time is too precious to even end professionally. I also agree with other reviews here about the red flags. If your company is aiming for remote first single IC, why do you force pair programming as your filter. Specially in the age of AI, fiddling with a syntax you are not familiar with without any help is just a waste of time.
After the recruiter intake, I attended the technical interview round. The process was intimidating and I honestly don't understand why they designed their interview this way. Some bug had to be detected in their product (fine), however it had to happen through puppeteering the interviewer and telling them instructions, while asking them to skip back and forth between screens (editor, browser) and navigate an unknown codebase and product. It seemed to me like a very artificial scenario and I went away feeling placed in a position where success is impossible.