I applied through college or university. I interviewed at IBM
Interview
I applied on campus career portal with link to the official website of IBM. Then went through a 90 minutes on-demand interview with 13 technical questions asking me some concepts of statistics and machine learning. Not very difficult. Then I had a phone interview that asked me some past experience in data science and some typical interview questions such as why IBM, why this position and so on.
Interview questions [1]
Question 1
Why do you think your background is a good fit for IBM
I applied through a recruiter. The process took 2 months. I interviewed at IBM (Austin, TX) in Mar 2016
Interview
Several conversations to pre-screen -- one was UK contact, one was email rapport and one was the a recruiter from San Diego. Then an hour-long interview with a senior manager about the role/roles, the part of the company and some of the exciting new directions and strategy. Then a technical interview by phone with a future colleague.
Interview questions [1]
Question 1
Technical: Topics on machine learning I have done in the past such as why I used a Monte Carlo to help with stress testing an algorithm for FDA approval or what programming language I used for creating an ensemble algorithm. The interviewer was a future colleague and extremely sharp.
Managerial: mainly my listening and asking questions only to understand better.
I applied through an employee referral. I interviewed at IBM (Los Angeles, CA) in Mar 2016
Interview
I was referred to the IBM Watson Health Group - Innovation team by a previous coworker who works for IBM now. I talked with the HR on the phone. It is an information interview for me, but certainly she took it as a screening. I heard that she was typing while I talked about my research. She then asked a data scientist to give me a technical phone interview. The technique interview was not too hard, but certainly you need to prepare for it because she asked about the detailed questions on data modeling.
Interview questions [2]
Question 1
Q. What is the matrix used to evaluate the predictive model?