YouTube

Role: Software Engineer Intern

College: University of California, Berkeley

Interview Date: January 2024

Interview Type: Internship

Result: Rejected

Difficulty: Medium

Rounds: 3 rounds

Topics Asked: Data Structures (Arrays, Strings), Algorithms (Sorting, Searching), Object-Oriented Design.

Experience:

This candidate shared their experience interviewing for a YouTube Software Engineer Internship. Despite being rejected, the candidate found the process insightful. The focus was on foundational CS concepts and coding ability for an intern role.

Round 1: Online Assessment (90 mins)

Two coding questions on HackerRank. One was a string manipulation problem, something like "find the longest palindromic substring." The other involved array manipulation and sorting, similar to "merge k sorted lists" but with simpler constraints. The candidate solved one completely and partially solved the other.

Round 2: Technical Phone Interview (45 mins)

Interviewer was a current YouTube engineer. Started with behavioral questions about the candidate's projects and why YouTube. Then, a coding question: "Implement a basic LRU Cache." The candidate explained the use of a `HashMap` and a `DoublyLinkedList` for O(1) operations and coded it up, correcting a few edge cases with hints.

Round 3: Virtual Onsite (2 x 45 mins)

Two back-to-back technical interviews.
Interview 1: Mostly behavioral and some object-oriented design discussion. "Design a class hierarchy for different types of YouTube videos (e.g., short, long, live stream)." Discussed inheritance, polymorphism, and relevant attributes/methods.
Interview 2: Another coding question. "Given a 2D grid representing a map of '1's (land) and '0's (water), count the number of islands." This was a standard BFS/DFS problem, and the candidate implemented BFS. The interviewer asked about optimizations and handling very large grids.

Overall experience: The interviewers were friendly and helpful. The candidate felt their OOD and BFS solutions were decent, but perhaps the initial struggle with LRU and one OA problem cost them the offer. They suggested practicing more on speed and perfecting solutions under pressure.