I applied online. The process took 2 months. I interviewed at Recursive (Tokyo) in Jul 2024
Interview
There are in total 4 interview rounds.
The first is a typical Python coding test where you have to solve three (easy - medium) Leetcode questions.
The second is a Machine Learning specific interview, where you are asked basic questions regarding (modern) Machine Learning, i.e. Gradient Descent, Neural Networks, Optimizer etc. For that the interviewer discusses a take-home assignment that you have to submit prior to the interviews.
The third is a 30-minute presentation of a recent AI paper of your choice. Afterwards, there is a discussion with the audience where you have to answer questions. The questions might also be somewhat unrelated to the presentation itself.
The fourth is directly after the presentation, which is a short "culture screen". After the interviewer explains a little bit in more detail what the company does, he asks for your desired wage, your desired location (Japan/Abroad), whether you need sponsoring etc. In the end he also asks a question regarding your personality.
Interview questions [1]
Question 1
Tell me about a moment in your life where you utterly failed.
I applied online. The process took 2 weeks. I interviewed at Recursive in May 2025
Interview
1. Homework on Machine Learning (Four questions related to machine learning, modeling, and optimization)
2. Leetcode interview (Recruitment was actively stopped)
3. Homework Review Call
4. Paper Presentation + Cultural Call
5. Final Interview
I applied online. The process took 6 weeks. I interviewed at Recursive in Sept 2024
Interview
The interview process is quite lengthy.
This is primarily because the next available time slot for interviews is often more than two weeks away.
I didn’t pass the homework assignment review stage.
In that interview, candidates are typically asked about general machine learning knowledge, with the expectation of specific answers.
Interview questions [1]
Question 1
General machine learning knowledges.
SGD, hyperparameters that will affect training results, difference between classification loss and regression loss, etc.