Meta Interview Questions — Page 230

Meta·Software Engineer·Technical Phone Screen

Jul 2026

Meta SRE interview focused entirely on production operations, incident response, and server fundamentals. No coding, no LeetCode, just a long scenario-based conversation about what you'd actually do if something broke on your watch. Felt more like a war-story session than a traditional interview.

You're the on-call engineer for a production server for several days. How do you make sure it keeps running normally while it's your responsibility?
Walk me through how you'd handle an incident end to end, from detection to resolution.
User requests are hanging or returning too slowly. What are the likely causes and how do you isolate them?
How would you quickly figure out which programs and services are currently running on an unfamiliar server?
What are the key finite resources on a server that can become bottlenecks?
If you were advising a non-technical person running a website or small service, what operational best practices would you tell them to make the system easier to monitor and maintain?

“I started talking about dashboards and alerting thresholds and then realized I was basically describing reactive monitoring, not proactive ownership.”

View Post

Meta·Data Scientist·Onsite - Multi Round

Jul 2026

Brutal Meta DS onsite. One massive multi-part causal inference question covering everything from metric definition to Rosenbaum bounds. The kind of question where you realize halfway through that you've been answering a different version of it in your head.

Among Meta Quest users, those who use social features are hypothesized to be more regularly engaged than those who use game features. Given ~8 weeks of activity data, design a full analysis plan: define your primary engagement metric and guardrails, specify a randomized experiment if possible, propose an observational fallback with causal inference methods, compute the required sample size for a two-proportion z-test (baseline p0=0.35, MDE=+3pp, alpha=0.05, power=0.80), describe your estimator and how you handle multiple comparisons, detail at least five validity threats and robustness checks, and define a decision rule plus a communication plan for stakeholders.

“This was basically seven questions stapled together and presented as one.”

View Post

Meta·Software Engineer·Onsite - Behavioral / Leadership

Jul 2026

Behavioral round at Meta for a software engineering role. Three questions, all focused on how you've worked with other teams and how you've handled things going sideways. Pretty standard for this type of round but they go deeper than you'd expect.

Tell me about a time you worked across teams on a project. What conflicts came up, and how did you get everyone aligned on goals, timelines, and who owned what?
Describe a project that didn't go well or fell short of expectations. What was your role and what actually went wrong?
Looking back at that failure, what did you actually learn and what would you change about how you approached it, whether that's the process, the technical decisions, or how you managed stakeholders?

“This one I felt okay about but in hindsight I spent too long on the conflict part and not enough on how we actually resolved it.”

View Post

Meta·Data Scientist·Onsite - Behavioral / Leadership

Jul 2026

Meta onsite behavioral loop for a Data Scientist role, senior IC level. The whole round is built around one anchor question about your hardest recent project, then the interviewer peels back layers on influence, people, and growth. Came away thinking I underprepared the stakeholder and cross-functional angles relative to the technical stuff.

Walk me through the most technically or organizationally difficult project you've worked on in the past year. What was the goal, what made it hard, and what did you actually do?
Tell me about the hardest stakeholder you had to win over on a project. What were their objections and how did you handle it when they still pushed back after you showed them data?
Describe a time you collaborated with people who had very different work styles or backgrounds. What friction came up and how did you get past it?
Give me a concrete example of how you helped a teammate succeed, whether that was coaching, unblocking them, or making sure they got credit. What was the measurable result?
A new person joins your team. How do you actually onboard them, covering both the technical ramp-up and the unwritten culture and process stuff?
How do you build working relationships with cross-functional partners like Product, Engineering, or Policy? Tell me about a specific time there was real tension and how you resolved it.
Tell me about something important you had to learn fast for a project. How did you ramp up and how did you know you'd actually learned the right thing?
Quantify the business impact of your work and explain the trade-offs you accepted to get there, like accuracy versus latency, or moving fast versus doing it right.
Looking back at that project, what would you do differently? Walk me through the specific decisions you'd change.

“This is the anchor for the whole loop, so everything else branches off it.”

View Post

Meta·Data Scientist·Technical Phone Screen

Jul 2026

Meta DS interview with a gnarly causal inference question about a feature that only auto-loads for opted-in power users, so you can't just run a clean A/B. The whole thing was a deep dive into quasi-experimental methods and how to communicate bias risk to stakeholders.

A feature is auto-enabled only for power users who opted in, so you can't randomize cleanly. Which quasi-experimental method would you use (DiD, Synthetic Control, IV, or PSM), and why? Walk through the identifying assumptions for your choice.
How would you design placebo tests and pre-trend checks for your chosen method, and what robustness checks would you run?
What sensitivity analyses would you run, and how would you use negative controls or Rosenbaum bounds?
How would you explain the directional bias risks of this quasi-experimental approach compared to a proper randomized experiment, to a non-technical stakeholder?

“This is where I spent most of my mental energy.”

View Post

Meta·Data Scientist·Technical Phone Screen

Jul 2026

Meta data scientist interview with a pretty involved stats/ML question around Bayesian classification. The whole thing centered on a single meaty problem, which I wasn't expecting to go as deep as it did.

You have a population of reviewers and limited labels. Propose a classification rule using Bayes' theorem to decide if a reviewer is lazy versus careful, derive the false-positive and false-negative rates for that rule, and then explain how those error rates change if every reviewer is required to complete a large fixed number of reviews.

“Took me a minute to get my footing.”

View Post

Meta·Data Scientist·Technical Phone Screen

Jul 2026

Meta DS interview with a deep-dive experiment design question on ad insertion in the main feed. Single question but it covered basically every angle of A/B testing you can imagine, from randomization unit to long-term holdback to network spillovers. Dense.

You're planning to insert one extra ad every 8 organic posts in the main feed. Walk through the full experiment design: randomization unit and rationale, primary and guardrail metrics, power analysis for a +2% revenue MDE, SRM checks and novelty effect detection, ramp strategy with geographic and age holdouts, a long-term holdback for delayed churn, and how you'd handle network spillovers and bias from dynamic ranking reacting to treatment.

“This one sprawled in a way I wasn't fully ready for.”

View Post

Meta·Data Scientist·Technical Phone Screen

Jul 2026

Meta DS interview with a SQL question focused on Oculus app engagement. Pretty standard analytics prompt but the NULL handling detail tripped me up a bit.

Given a user activity table and an apps table, compute the percentage of total session time spent in each app category over the last 7 days relative to a given reference date. Handle cases where an app_id has no matching category.

“The core math is straightforward, a simple ratio of category duration to total duration.”

View Post

Meta·Software Engineer·Technical Phone Screen

Jul 2026

Meta coding screen for a software engineering role, focused entirely on designing and implementing a production-ready FastAPI service from scratch. No algorithms, no LeetCode, just real backend structure and clean code.

Design and implement a minimal but production-ready FastAPI service for text summarization, including a POST /summarize endpoint, Pydantic request/response models, in-memory storage of request/response pairs, and a clean separation of router, service, and model layers.

“This was the whole interview, basically.”

View Post

Meta·Data Scientist·Technical Phone Screen

Jul 2026

Meta data scientist SQL round, one question about conditional aggregation for a shop-level analytics metric. Pretty focused, no fluff.

Given a table of shop events with impression and view event types, write a SQL query that returns daily impressions, views, and a visibility ratio (views divided by impressions) for each shop, ordered by date and shop ID.

“The core of the question is conditional aggregation, which I knew, but I fumbled a bit deciding between FILTER and CASE WHEN syntax.”

View Post