Meta interview write-ups can be frustratingly brief about the actual problems. This guide keeps the useful distinction: one phone-screen account names its prompts, while an E4 candidate describes the round difficulty but leaves the problem titles out. Unnamed questions stay unnamed here.
Meta SWE phone screen
A candidate described an easy and a medium question in the same phone screen. The first was Max Top K Elements in an array, solved with a min-heap. The second asked for a class with get, put, delete, and last operations, where last returns the most recently accessed non-deleted key. The author discussed several data-structure choices and reported an O(n) target for the operations. Both prompts come from the Meta phone-screen report on LeetCode Discuss (2025).
The account also mentions a rough start caused by a webcam problem. That detail is practical: test the equipment and keep a backup ready before the call, so a setup issue does not consume problem-solving time.
Meta E4: what the source does and does not say
An E4 candidate described two interviews, each with a pair of problems that moved from easy or medium toward harder follow-ups. The candidate did not publish the problem names. The Meta E4 interview report (2025) therefore supports a note about pacing and problem-reading, but it cannot support a named question list.
The candidate's advice was to slow down, read the prompt aloud, restate the task, and explain an approach before coding. Those are reflections from one interview, not official Meta scoring criteria.
How to prepare for the Meta phone-screen and E4 reports
For the top-k prompt, compare a heap with sorting and be ready to explain how the cost changes with input size. For the key-value interface, clarify whether operations need to be constant time before choosing a list, map, or other structure. The original phone-screen post says the proposed operations were expected to run in O(n), so do not silently replace that requirement with a different one.
This is a small source set. It does not establish Meta's current loop or prove that these prompts recur. Use the interview question guide for practice structure and the company interview hub to compare other firsthand accounts.
Additional Meta candidate reports
E4 follow-ups and behavioral questions
A separate E4 report records two follow-ups to a binary-range problem, followed by three behavioral questions. The captured notes do not include the initial coding prompt, so only the reported follow-ups are listed:
- What changes if the digits range from 0 to 9 instead of being binary? (LeetCode Meta E4 report)
- How should the solution handle
start > end? (same report) - Tell me about a project you are most proud of and your role in it. (same report)
- Tell me about a production bug or something that went wrong because of you. What did you learn? (same report)
- What do you do outside work to learn or improve? (same report)
Report with no stated role or level
Another candidate described a 15-minute task to print a binary tree in a specific format. The post title does not identify the role or level, so this is kept separate from the E4 and phone-screen reports. (LeetCode Meta report)
Frequently asked questions
Does the Meta E4 report identify the coding problems?
No. That candidate described the difficulty and progression of two interviews but did not publish problem names. This page therefore reports the account without reconstructing or guessing the questions.
Run it on Parlel
When you have chosen a target team, use the role's actual scope to guide your preparation. Parlel's company directory can help you map the organizations you want to contact.