Level: Senior-Level
Round: Full Journey · Type: Multiple Types · Difficulty: 6/10 · Duration: 300 min · Interviewer: Unfriendly
Topics: Data Structures, Algorithms, System Design, Behavioral Questions, Binary Tree, ML/AI Infrastructure
Location: San Francisco Bay Area
Interview date: 2026-03-13
Got offer: False
Here's my recent interview experience at Meta.
Background: I have 7 years of experience, primarily in ML/AI infrastructure. Previously, I worked at a large company on data/ML infrastructure. Recently, I've been working on LLM agent/ML infrastructure/distributed systems. I interviewed for a Meta E5 SWE role focused on AI/ML infrastructure.
2/25: Phone Screen: The first question was too easy to remember. The second question was left + right side view of a binary tree. I was notified the next day that I passed and scheduled the onsite interview.
3/9 - 3/13: Onsite:
Coding: I solved two problems quickly and bug-free in 25 minutes. The problems were "LeetCode 403" and "LeetCode 227". The interview ended early.
Behavioral: Standard behavioral questions.
AI Coding: The question was a maze-type problem with approximately 7 test cases. The interviewer only asked me to complete the first 4 test cases. There was a small issue: the interviewer initially wrote the wrong expected output for the first test case but apologized after realizing the mistake. I completed all 4 test cases before the end of the interview.
System Design: The question was to design LeetCode. I answered all the interviewer's questions smoothly.
3/16: HR called to say no hire.
LeetCode similar: LeetCode 403, LeetCode 227