← Back to Directory

Meta interview notes, page 132 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·Machine Learning Engineer·Onsite - System Design / Architecture

May 2026
Meta ML engineer interview, system design round focused on recommendation engines. Pretty open-ended prompt with not much scaffolding given.
  • Design a recommendation engine from scratch.

“Spent the first few minutes just trying to scope it down because 'recommendation engine' could mean a hundred different things.”

View Post

Meta·Software Engineer·Technical Phone Screen

May 2026
Meta data engineer interview, SQL-heavy technical screen that ran in two parts. Part B was the one that actually stressed me out: you get the query results and have to find what's wrong with the data, no tools allowed.
  • Given the output of a SQL query, identify all data quality issues present in the results, such as negative values in fields that should never be negative, unexpected NULLs, duplicate rows from a join, or timestamps outside a valid range.
  • After identifying the data quality issues, update the query to handle them correctly using filters, COALESCE, deduplication logic, or validation checks, and explain how you would surface these issues in a production environment.

“This was the part I underestimated.”

View Post

Meta·Machine Learning Engineer·Technical Phone Screen

May 2026
Meta MLE debug round, no AI tools allowed. One focused problem about a broken recommendation function. Pretty contained scope but the constraint of no assistance made it feel higher stakes than it probably was.
  • A `valid_recommend(user, candidates)` function is supposed to validate a friend recommendation list for a given user, but an existing test is failing. Find the bug and fix it with a minimal, targeted change.

“The fix itself is small, like a one-liner, but I spent way too long second-guessing myself because I kept looking for something more complex.”

View Post

Meta·Product Manager·Onsite - Product Sense / Strategy

May 2026
PM interview at Meta with a single estimation question. Pretty standard case-style format, nothing too wild, but these market sizing questions always make me second-guess my assumptions mid-answer.
  • How many flights depart London for Dubai on a typical day?

“I started with airport capacity which in hindsight was the wrong entry point.”

View Post

Meta·Software Engineer·Technical Phone Screen

May 2026
Meta software engineering interview with two coding problems back to back. Nothing too exotic but the second one had a wrinkle I didn't see coming at first.
  • Given an m×n binary grid where 1s are blocked cells and 0s are passable, determine whether a path exists from the top-left to the bottom-right corner using up/down/left/right movement.
  • Given an integer array and a value k, return the k most frequently occurring numbers. Ties in frequency can be broken arbitrarily.

“Went with BFS pretty quickly, felt solid.”

View Post

Meta·Software Engineer·Technical Phone Screen

May 2026
Interviewed for a data engineer role at Meta. The content was pretty thin so I'm not entirely sure what round this was, but it seemed to be centered around product analytics concepts using Instagram Likes as the core case.
  • How would you define and measure the success of Instagram Likes as a product feature?

“This is a classic Meta-style metrics question and I probably overthought it.”

View Post

Meta·Data Scientist·Technical Phone Screen

May 2026
Meta DS interview, one big meaty case study on metric drops. The whole thing was basically a single multi-part question that kept going deeper every time I thought I'd answered it. Walked out unsure if I'd nailed it or just talked a lot.
  • A key metric (comments per daily active user) drops significantly. How do you scope and sanity-check the drop before diving into root cause analysis?
  • Walk through how you'd segment the data to localize where the drop is coming from, and explain how you'd distinguish a mix-shift from a genuine within-segment decline.
  • What internal and external causes would you consider for a metric drop like this, and how would you validate or rule out each one?
  • Which time-series or causal inference methods would you use to detect and attribute the anomaly, and how would you set up controls to avoid contamination from the treatment period?
  • How do you decide when you've actually found the root cause versus just a correlation, and what does your rollback and monitoring plan look like afterward?

“My instinct was to jump straight to segmentation but they wanted me to slow down first.”

View Post

Meta·Software Engineer·Technical Phone Screen

May 2026
Meta SWE coding round with a simulation-style problem about battery usage. Pretty straightforward premise but the edge cases kept me second-guessing myself.
  • Given a total usage time, a list of battery capacities, and a list of recharge times, determine how many consecutive batteries are needed to cover the full usage time. Return -1 if all batteries are still recharging.

“The problem sounds simple on the surface but the recharge timing tripped me up.”

View Post

Meta·Software Engineer·Onsite - Behavioral / Leadership

May 2026
Behavioral screen for a data engineer role at Meta. Just the one question from what I can tell, pretty standard stuff.
  • Tell me about a mistake you made and what you learned from it.

“Picked a mistake that felt safe and ended up sounding like I was humble-bragging about a minor thing.”

View Post

Meta·Software Engineer·Onsite - System Design / Architecture

May 2026
Meta EM interview with a single system design question. Pretty open-ended, which I wasn't fully ready for.
  • Design a vending machine.

“Seemed deceptively simple at first and I think I wasted the first few minutes going too shallow.”

View Post