Meta Interview Questions — Page 83

Meta·Software Engineer·Technical Phone Screen

May 2026

Meta Research Scientist interview that was essentially a coding exercise around building an agent loop from scratch. One meaty implementation question, no fluff, and they clearly wanted to see how you think through control flow and error handling under pressure.

Implement a ReAct-style agent loop using a provided set of APIs and helper functions. The loop should cycle through reasoning, action selection, tool invocation, and result observation until a final answer is reached. You also need to handle termination conditions and error cases.

“This one took me a minute to even parse what they were asking.”

View Post

Meta·Software Engineer·Onsite - Behavioral / Leadership

May 2026

Behavioral round at Meta for a software engineering role, focused almost entirely on cross-functional collaboration. One meaty question that took up most of the time.

Tell me about a project where you had to work across team boundaries, including different orgs, reporting chains, geographies, or non-engineering stakeholders like PMs, designers, or legal. Walk through the goal, who the stakeholders were and where their priorities conflicted, how you got everyone aligned, how you handled disagreements, and what the concrete outcome was. What did you learn about influencing without authority, and what would you do differently?

“This one is deceptively long and I think I spent too much time on the setup and ran out of steam before getting to the actual disagreement part, which is clearly what they care about.”

View Post

Meta·Software Engineer·Online Assessment (OA)

May 2026

Straightforward Meta SWE screen, just a basic array sum problem. Nothing tricky, pretty much a warmup question.

Given an array of integers, write a function that returns the sum of all its elements.

“Pretty much as vanilla as it gets.”

View Post

Meta·Software Engineer·Technical Phone Screen

May 2026

Meta coding screen, pretty much just this one problem the whole time. Fast exponentiation sounds straightforward until you're actually on the clock trying to remember the edge cases.

Implement a power function that computes x raised to the n, handling negative exponents and the full 32-bit integer range for n, with better than linear time complexity.

“I knew the recursive halving approach going in but fumbled the negative exponent case for longer than I'd like to admit.”

View Post

Meta·Software Engineer·Technical Phone Screen

May 2026

Meta SWE interview with a tree-based algorithm problem. Not the hardest thing I've seen but I definitely overthought the traversal part before settling on something that actually worked.

Given a Binary Search Tree, find the length of the longest increasing subsequence while preserving the original node order in the tree.

“My first instinct was to just do an inorder traversal and run a standard LIS on the result, which would've been wrong because inorder on a BST already gives you sorted values and that's not the same as preserving node order.”

View Post

Meta·Software Engineer·Onsite - Coding / Algorithms

May 2026

Meta SWE interview with three back-to-back coding tasks that mixed classic algorithms with from-scratch ML implementation. No ML libraries allowed, which I was not expecting going in.

Given a string, find the length of the longest contiguous substring where all characters are distinct.
Implement binary logistic regression from scratch, including fit, predict_proba, and predict methods, using only batch gradient descent and no ML libraries.
Implement a multinomial Naive Bayes text classifier from scratch, including Laplace smoothing, log-space computation, and handling unknown tokens at test time.

“Classic sliding window.”

View Post

Meta·Data Scientist·Onsite - Product Sense / Strategy

May 2026

Product feature scoping question at Meta for a DS role. Five-part case covering everything from pre-launch signal analysis to novelty effect detection. Dense but pretty representative of what they actually care about.

Using only pre-launch observational data, how would you determine whether users actually need a proposed new in-app sharing feature? What signals would you look for and how would you segment to avoid drawing misleading conclusions?
If you could instrument or collect any data you wanted, what would you add to better understand whether users need this feature? How would you combine qualitative and quantitative inputs?
After the feature launches, how would you measure success? Define a North Star metric and a few supporting or guardrail metrics, and walk through how you'd design the A/B test.
How would you work with stakeholders to clarify what the feature is actually supposed to achieve and surface the key trade-offs before committing to a measurement plan?
How would you detect a novelty effect in your experiment results, and what would you do about it to make sure any observed lift is durable?

“This is where I probably over-indexed on engagement signals and forgot to define 'need' before diving in.”

View Post

Meta·Data Scientist·Onsite - Behavioral / Leadership

May 2026

Meta data scientist onsite with three pretty dense sections covering cross-team conflict resolution, rapid domain learning under pressure, and pushing back on a senior stakeholder's analytic choices. The questions were specific and situational, not the usual behavioral fluff, which I wasn't fully prepared for.

A peer on another team tells you they feel excluded from discussions that affect shared work, and influential members of the other team are driving that exclusion. You have no formal authority and a launch is three weeks out. Walk through how you'd diagnose the situation in the first 48 hours, including what questions you'd ask, what artifacts you'd look at, and what signals you'd watch for.
For that same cross-team conflict scenario, lay out your full stakeholder map and a concrete communication plan, including who you'd talk to, when, through what channel, and what you're trying to accomplish with each conversation.
Still in the same scenario: what immediate steps would you take to protect both psychological safety and your delivery timeline, given the launch pressure?
In that same conflict situation, what specific criteria would you use to decide it's time to escalate to HR or senior leadership rather than continuing to handle it yourself?
How would you define leading and lagging metrics to know whether your intervention in that conflict situation is actually working? Include both quantitative and qualitative measures, and explain how you'd track them.
What are two ways your intervention plan for that conflict could fail, and how would you adjust if you saw those failure modes emerging?
Tell me about a specific time you had two weeks or less to learn a new domain or tool in order to deliver something high-stakes. What did you cut, how did you confirm you were learning the right things, and how did you measure the impact afterward? Include a mistake you made.
Describe a real conflict you had with a senior stakeholder who disagreed with your analytic approach. How did you figure out what they actually cared about versus what position they were taking, what trade-offs did you put on the table, and how did you quantify the impact of whatever decision you ended up making?

“This tripped me up more than I expected.”

View Post

Meta·Data Scientist·Onsite - Product Sense / Strategy

May 2026

Product sense interview with a Head of Product at Meta for a Data Scientist role. One deep-dive project question that covered basically everything: metrics, experimentation, stakeholder drama, and what you'd change. Left feeling like I either nailed it or completely missed what they were looking for.

Walk me through a product or marketplace project where you had meaningful impact, covering the business problem, the metrics you chose, any analysis or experiments you ran, your recommendation, the actual outcome, and how you navigated ambiguity or pushback along the way.

“This is basically a 20-minute monologue disguised as one question.”

View Post

Meta·Product Manager·Onsite - Product Sense / Strategy

May 2026

Meta PM interview, product strategy round. Just one question but it's the kind that opens up into a rabbit hole fast if you're not careful about scoping.

You're a PM at Wikipedia. What product should you build next?

“This one tripped me up because Wikipedia feels so static that my brain defaulted to 'improve the editor experience' which is probably the most boring answer you could give.”

View Post