Snap Inc. · Onsite
Rejected2026 Snap Experienced Machine Learning Engineer Virtual Onsite: 1 Experience
Coverage: One experienced machine learning engineer with a Master's degree, changing jobs, who came to Snap through a recruiter. The report covers a four-round virtual onsite. It gives July to September 2026 as the interview period and went up on November 12, 2026.
What came up
Two of the four rounds were about recommender systems, one leaning toward retrieval and one toward ranking. The coding round was billed as AI coding but turned out to be a classic graph traversal question.
| Topic | Reports | Most recent report |
|---|---|---|
| Recommender systems: retrieval, ranking and ads labels (two rounds) | 1 of 1 | Nov 12, 2026 (2026 Q3) |
| Coding: graph traversal over parent-child relations | 1 of 1 | Nov 12, 2026 (2026 Q3) |
| ML fundamentals through a résumé deep dive | 1 of 1 | Nov 12, 2026 (2026 Q3) |
Questions we found
Algorithms / Data Structures
| Question | Frequency | All last seen dates |
|---|---|---|
| Given parent-child relations, traverse the graph under fairly involved conditions (the candidate used BFS for one level, then DFS) | 1 report | Report: 2026-11-12 (interview 2026 Q3) |
ML System Design
| Question | Frequency | All last seen dates |
|---|---|---|
| Recommender system design (Design a Personalized Recommendation System; the practice version has no ads part), asked in two rounds: one focused on retrieval and one on ranking, with ads questions such as how labels are computed | 1 report | Report: 2026-11-12 (interview 2026 Q3); additional FastPrep history: 2026-10-30 |
1 detailed candidate timeline, loops, and screens
01: MLE virtual onsite, two recommender rounds
Interview period: 2026 Q3 · Report date: November 12, 2026 · Role: Machine learning engineer · Degree listed: Master's · Channel: Recruiter · Candidate's rating: Neutral, average difficulty · Result: Not selected
Coding
- The round was called AI coding, but the candidate says there was effectively no AI part. It was a classic LeetCode-style question.
- The task gave some parent-child relations and asked for a graph traversal. The conditions were fairly involved, and the candidate's approach was a BFS over one level followed by DFS. The report doesn't spell out the conditions or the expected output, so this question has no practice link.
ML Design and Applied ML
- Both rounds were about recommender systems. In theory, one goes deeper on system details and the other on ML modeling details, but the candidate found little difference: both asked about a bit of everything.
- One round focused on retrieval and the other on ranking. One of them dug into ads details, such as how labels are computed.
- FastPrep's practice version, which lists Snap Inc. among its companies, is Design a Personalized Recommendation System. It covers retrieval and ranking for a general user-item catalog and has no ads part. The report gives no specific product or scale for these rounds.
ML Fundamentals
- This round followed the candidate's résumé, with a deep dive into it. The candidate rates it the best round of the loop, with many interesting detailed questions.
Result
The candidate was not selected.
Keep practicing
- System design: Design a Personalized Recommendation System covers the retrieval and ranking side of the two recommender rounds. More Snap design questions are in the Snap Inc. system design bank.
- Coding: more Snap questions, in every practice format, are in the Snap Inc. question bank.
- Everything Snap in one place: the Snap interview page and Snap Inc. hiring insights.
- More firsthand reports: browse the Experience feed.
Sources: Anywhere useful interview info can be found.
Comments
0No comments yet — be the first to share what you know.