Goldman Sachs

I went through the Goldman Sachs Summer Analyst hiring process through the campus drive. The first stage was an online assessment, and it was not purely a coding test. I had to manage three DSA questions along with aptitude and a situation-based question. The coding problems were the part I prioritized because I knew that spending too much time on one difficult question could hurt my performance in the other sections. I was able to clear the assessment and move to the interviews. Recent campus reports describe a similar combination of three coding questions, aptitude and a situation-based essay. My first interview was fairly fast-paced. After a short introduction, the interviewer asked about one of my projects and started digging into the implementation. I was asked why I had selected particular technologies and whether I had considered other approaches. The coding portion then started. One question was about the difference between Merge Sort and Quick Sort. I was asked to explain the recursion and complexity before writing Merge Sort on paper. I had to walk through the recursive calls and explain the best, average and worst-case behavior. The interviewer then moved to linked lists and asked about cycle detection. I explained Floyd's slow-and-fast pointer technique and discussed why it works. There were also a few SQL and operating-system questions. I noticed that the interviewer wasn't satisfied with only definitions; whenever I gave an answer, there was usually a small follow-up asking me to apply the concept. The second round focused much more heavily on my project. We discussed authentication and authorization, JWTs, access tokens and refresh tokens. One scenario involved a stolen token, and I had to explain how I would reduce the security impact instead of simply saying that JWTs are secure. This led to a fairly detailed conversation about how my application handled authentication. At the end of the round, I was given a DSA problem: reverse a linked list in groups of K. I had to write the complete solution on paper within a limited amount of time. I knew the standard approach, but keeping track of the pointers while explaining the code was challenging. A recent Goldman Sachs campus report describes the same type of second-round project discussion and K-group linked-list problem. The final stage was shorter and more conversational. I was asked about my interest in software engineering, teamwork, and situations where I had taken ownership of something. There were also some logical questions. I focused on keeping my answers specific and connected to things I had actually worked on. After completing the final round, I received the selection confirmation. The process made me realize that Goldman Sachs interviews can move between DSA, core CS, project architecture and behavioral questions very quickly. Being comfortable with only one of those areas would not have been enough for me.