Expedia SDE-3 GenAI/ML — elimination coding round (pending)

Sharing my interview experience from today for an SDE3 role focused on GenAI/ML at Expedia. This was a 1-hour elimination round. Focus: DSA + problem-solving + some discussion around approach. No direct deep ML questions in this round.

Problem 1: Merge Overlapping Intervals
Classic interval merging. I initially approached it using a two-pointer style comparison between adjacent intervals, realized mid-way something felt off with maintaining merged state, discussed sorting and overlap conditions but didn't fully converge to the clean greedy solution. Optimal: sort intervals, iterate and merge using the last added interval.

Problem 2: Number of Islands
Standard DFS/BFS grid traversal. I used DFS, covered edge cases (visited marking, bounds), code ran correctly.

The interviewer was neutral and let me drive the solutions, with some probing questions around approach and correctness.