← Back to Directory

Meta interview notes, page 100 of 238

FAANG / Big Tech

Most of these Meta write-ups come from Software Engineers (1,194), then Data Scientists (489) and Product Managers (453). The most reported rounds are the technical phone screen (1,190), the product sense round (466), the system design round (255) and the behavioral / leadership round (205).

2377 interview notes · updated Jul 2026

Meta·Product Manager·Onsite - Product Sense / Strategy

May 2026
Interviewed for a PM role at Meta and got hit with a product metrics question about Zoom. Pretty standard stuff for a product sense round but I fumbled it more than I expected.
  • What key metrics would you use to measure the success of Zoom?

“I went straight to DAU and meeting minutes which felt obvious the second I said it out loud.”

View Post

Meta·Data Scientist·Technical Phone Screen

May 2026
A technical screen for a Data Scientist role at Meta that went deep fast. The whole session was basically one extended problem about call quality modeling, and it had enough sub-parts that it felt like a mini take-home crammed into a live interview.
  • You need to set a maximum participant cap K for group video calls. Given call-level telemetry (participants, quality scores, jitter, packet loss, etc.), propose a statistical method to choose K that maximizes total successful participant-minutes while keeping P95 median MOS score at or above 3.8 and dropped call rate at or below 2%. Walk through: (a) fitting and validating a change-point or segmented regression model of call quality vs participant count, (b) computing uncertainty on K and reporting a 95% CI within plus or minus one participant, (c) whether to output a single global K or dynamic caps by device and network type and when to collapse to a single limit, and (d) how much data you need and how you'd handle heavy-tail outliers.

“This is a beast of a question.”

View Post

Meta·Data Scientist·Technical Phone Screen

May 2026
SQL-heavy technical screen for a Data Scientist role at Meta, centered on an e-commerce scenario with joins, aggregations, and date filtering. Nothing too exotic but the compound nature of the second question tripped me up a bit.
  • Given an interactions table and a products table, write a SQL query to find how many products have more than 3 distinct buyers AND more than 10 total interaction units. Then, as a follow-up, calculate the percentage of 'validate' category interactions for U.S. products within the last 7 days.

“Two questions bundled into one and I didn't clock that fast enough.”

View Post

Meta·Software Engineer·Technical Phone Screen

May 2026
One coding round at Meta for an infrastructure role, two problems back to back. Nothing too wild but the second one had more edge cases than I expected.
  • Given two nodes in a binary tree where each node has a parent pointer but there's no reference to the root, find their lowest common ancestor.
  • Given a string containing parentheses and lowercase letters, remove the minimum number of parentheses to make the string valid.

“The trick is you don't have the root, so you can't do the usual recursive thing.”

View Post

Meta·Machine Learning Engineer·Technical Phone Screen

May 2026
Meta ML engineer coding round, one problem the whole session. Classic bitmask DP problem that I'd seen before but still managed to overthink in the moment.
  • Given a list of strings, find the maximum length of a concatenation of a chosen subset such that every character in the result is unique.

“Knew this problem.”

View Post

Meta·Machine Learning Engineer·Onsite - Coding / Algorithms

May 2026
Meta MLE interview built around a single maze problem that kept getting harder across four parts. The structure was clever but also kind of brutal if you hit a wall early, since each part builds directly on the last.
  • Implement a basic maze solver (no AI tools allowed): given a 2D grid with walls, open cells, a start, and a goal, determine if the goal is reachable and return a shortest path.
  • Extend the maze solver to return all shortest paths from start to goal, not just one.
  • Add a meaningful extension to the maze solver, such as weighted movement costs, teleporters, or multiple goals, and justify your algorithm choice among BFS, Dijkstra, and A*.
  • Tackle a harder variant of the maze problem, such as walls that change over time or batched queries. Discuss the time and space complexity, and explain how you'd validate correctness when AI wrote most of the code.

“The no-AI constraint on Q1 felt like a vibe check more than anything.”

View Post

Meta·Software Engineer·Technical Phone Screen

May 2026
Meta SWE coding round with a tree traversal problem. Pretty standard backtracking stuff but the negative numbers constraint is what trips people up if they're not careful.
  • Given a binary tree and a target sum, find all root-to-leaf paths where the sum of node values along the path equals the target. Return each valid path as a list of node values. The tree can contain negative numbers and up to 5000 nodes.

“The negative numbers part is the thing I almost missed.”

View Post

Meta·Data Scientist·Technical Phone Screen

May 2026
SQL-heavy technical screen for a Data Scientist role at Meta, focused on messaging platform analytics. Two questions, both involving joins and aggregations on user behavior data. Nothing too wild but the spammer read-rate one had a subtle denominator definition that I almost got wrong.
  • Given a messages table with sent and read dates, how would you find the number of users who received more than 10 messages in a single day during the past 7 days?
  • Using the messages and violating_scores tables, calculate the average read-rate for spammers over the last week, where read-rate is the number of messages sent by spammers that were read divided by all messages sent by spammers.

“Pretty standard aggregation question.”

View Post

Meta·Software Engineer·Onsite - Coding / Algorithms

May 2026
Meta coding round for a software engineer role. One question, but it had enough moving parts to keep me busy for the whole session.
  • Given the head of a singly linked list and a target value v, find the first two nodes whose values equal v (call them A and B, A appearing before B), then reverse the sublist strictly between A and B in place. If fewer than two such nodes exist, return the list unchanged. Can you do it in one pass with O(1) extra space?

“The core reversal wasn't the hard part.”

View Post

Meta·Software Engineer·Technical Phone Screen

May 2026
Meta SWE interview with a memory management question that sounds basic until you're actually in the room trying to remember what goes on the heap versus the stack.
  • Walk through the memory changes that happen when you execute: Object o = new Object().

“I knew the general answer but fumbled the sequencing.”

View Post