I applied online. I interviewed at Microsoft in Sep 2025
Interview
Screening interview basic ml question variance bias, dropout, basic leetcode question around easy and medium. Basicly explain what you do and how to execute it. No need to write code. Just explain the idea. Detail about your project.
The interview process started with a short 30-minute introductory interview, where we mainly discussed my previous experience, background, and motivation for the role. After that, I had three interviews in one day. One interview focused more on my previous projects and experience, while the other two were more technical and focused on machine learning theory, practical ML problem-solving, and model selection. The interviewers were very nice, professional, and the atmosphere was comfortable throughout the process.
Interview questions [2]
Question 1
Describe how you would build a model that recognizes whether a person is wearing glasses in a photo.
I applied online. I interviewed at Microsoft (Cambridge, MA) in Apr 2026
Interview
The interview process included an initial screening stage followed by a final round. The final round consisted of three 45-minute interviews with short breaks in between. The discussions focused mainly on prior experience, research projects, machine learning fundamentals, model evaluation, experimentation, and applied problem-solving in product-oriented settings. There were also behavioral questions and discussion of how I approach failure analysis, data quality issues, and the practical impact of technical solutions.
Interview questions [1]
Question 1
One time you had to modify an existing algorithm. Tell me about your research.
Two rounds of interview, both technical, in both rounds I was asked about my projects and very deep insights related to those, also the questions were very practical based ml questions, i was asked to do live coding also in both the rounds