I had applied to several positions at Intel and, tired of rejections, I was totally surprised by this opportunity. I received an email from a technical lead at Intel who wanted an informal interview to discuss my background and a bit of my research work. The interview was with the Advanced Visual Computing team.

The informal interview was relatively easy. The lead went over my resume, asked a couple of questions about my projects and research, and inquired if the position aligned with my interests. She also asked some behavioral questions to assess my fit at Intel. The interview went well, and she asked if she could schedule a technical interview for the following week over the phone to move the process forward.

The technical interview was average to hard. Although it was for a graphics intern position (Machine Learning and Computer Vision), I was asked questions on caching, L1 cache, and pipelining, along with some basic C questions (pointers, the extern keyword, debugging, etc.). She then transitioned to Matlab questions (basic to average) involving plotting, the rand function, and so on. I was asked 3-6 questions on image processing, including topics like adding noise, image registration, image filtering, and clustering. The interview concluded with a couple of basic machine learning questions (supervised learning, classification). The interviewer was very friendly and made me feel comfortable throughout. She mentioned she would get back to me soon with a decision. I heard from Intel corporate a week after the interviews.

Intel is flexible with the use of software languages. As a proficient Matlab user, I was given the flexibility to use Matlab over C++. This is why I felt very comfortable during the interviews.

What is a convolution filter and a Gaussian filter?

Explain why one is better than the other and its applications.