Amazon

I applied for the Amazon SDE Intern role through the off-campus process. The selection process started with an online assessment that had four different sections: debugging, coding, behavioral questions, and logical reasoning. The debugging section consisted of seven small code snippets where the objective was to make minimal changes so that the given code produced the expected output. The coding section had two medium-level problems, one based on maps and another involving graphs. There were also behavioral questions and a reasoning section containing aptitude and verbal ability questions. After clearing the assessment, I received the interview schedule. There were two technical rounds, and both were elimination rounds. Technical Interview 1 The first interview lasted around an hour. After a quick introduction, the interviewer moved directly to problem solving. The main question was a word-search type problem on a character board. I initially approached it using DFS and explained the traversal and complexity. The interviewer then asked whether the solution could be optimized further. After thinking through the problem, I discussed a Trie-based approach, which could avoid repeatedly exploring unnecessary paths. Once the approach was finalized, I was asked to implement it on the live coding platform. The interviewer was quite interactive throughout the discussion and kept asking why I was choosing a particular approach instead of simply looking for the final code. The round went slightly beyond the scheduled time, so there wasn't much time left for questions at the end. A few hours later, I received a call from HR informing me that I had cleared the first interview and that the second round would take place shortly. Technical Interview 2 The second round again started with a brief introduction. This time, I was first asked to explain one of my projects before moving into DSA. The first coding problem involved searching for an element in a sorted and rotated array. I initially described a straightforward approach, but the interviewer asked me to optimize it. I tried modifying binary search and discussed the different cases depending on which half of the array was sorted. The second problem was related to finding the left view of a binary tree. I explained the BFS approach, tracked the first node encountered at every level, and then implemented the solution. The interviewer also focused on complexity and asked me to walk through sample cases rather than immediately jumping into code. I completed the second problem, but the first question required quite a bit of guidance. After the interview, I was informed that I had not been selected. The process was challenging but useful because the interviews were heavily focused on how I approached unfamiliar problems. The interviewers were willing to provide hints, but I realized that being able to independently move from a brute-force solution to an optimized one was just as important as writing correct code.