I applied online. I interviewed at AT&T in Oct 2018
Interview
Applied online and received an email to set up a video interview. It was standard interview with a few questions about my resume and then proceeded to ask technical questions.
Interview questions [1]
Question 1
(1) Asked about the projects listed in my resume.
(2) Describe curse of dimensionality with an example
(3) What is bias-variance tradeoff?
(4) Have you used transfer learning to solve a problem in image classification
I applied online. I interviewed at AT&T in Mar 2017
Interview
It was a very good process. Received an email from recruiter for a telephonic conversation. After the conversation , technical round on phone in next 3 days and the final call from HR
Interview questions [1]
Question 1
Describe about the project and past work experience
I applied online. The process took 2 weeks. I interviewed at AT&T
Interview
First stage was a discussion around my background and the role with some technical data science questions, second stage was possibly the most pointless and total waste of time it was as if the interviewers were not interested in anything I had to say apart from asking me development questions rather than data science questions like a quick fire question match. The interviewer also had no clue how to ask questions properly as if they totally lacked experience or knowledge of the subject matter. Possibly, the worst phone interview I have had where I felt the interviewer had no clue about the data science process. Also, majority of their questions were unrelated to the role.
Interview questions [1]
Question 1
They asked me testing questions but most of those testing questions related to development not much on the model evaluation and benchmarking, nor about pipelining. Very pointless. Most questions were too simple and when I answered them from a pipeline or model evaluation process they would get lost. So tried to add clarity on the types of tests I would do in development as well as for model evaluation and across the different pipelining process to further add context to my answer and to their questions.