Snap Machine Learning Engineer Interview Questions

Snap·Machine Learning Engineer·Onsite - System Design / Architecture

Jul 2026

Snapchat MLE system design round, heavy on geo and ML pipeline stuff. The question was basically one giant beast that kept spawning sub-problems every time I thought I was done. Walked away feeling okay about the spatial indexing parts but less sure about how I handled the ANN vs. learned ranking discussion.

Design a real-time system that surfaces the top-K points of interest within a user's current map viewport as they pan and zoom, with p95 end-to-end latency under 150ms.
When would you use approximate nearest neighbor search for candidate retrieval versus a learned ranking model for re-ordering results, and how do you decide which to use?
How would you handle load shedding, deduplication across tile boundaries, and pagination while the user is actively panning?
How would you evaluate this system both offline and online, and what metrics would you track?
How do you handle cold start for new POIs or new users in this system?

“This one sprawled in every direction.”

View Post

Snap·Machine Learning Engineer·Technical Phone Screen

Jul 2026

Senior ML loop at Snapchat focused almost entirely on one deep-dive recommendation project. No LeetCode, just a long technical conversation about modeling decisions and tradeoffs. Left feeling like I'd either nailed it or completely exposed myself, still not sure which.

Walk me through the modeling approach you chose for a past recommendation or prediction project and explain why you made that choice.
Why didn't you go with a more complex model for that project?
How do you think about balancing precision, recall, latency, and infrastructure cost when making modeling decisions?
What would you do if your offline evaluation metrics look good but the online business metrics are moving in the wrong direction?
How would you go about tuning the decision threshold for a binary classifier in production?
Compare Factorization Machines with embedding-based deep neural networks for a recommendation or ranking task. When would you pick one over the other?

“I anchored on a two-tower retrieval model I'd built before and talked through why it made sense for the scale we were at.”

View Post

Snap·Machine Learning Engineer·Technical Phone Screen

Jul 2026

Snapchat ML Engineer coding round, two problems back to back. Both were more classic CS than anything ML-specific, which surprised me a little.

Build an arithmetic expression evaluator that parses a string with non-negative integers, spaces, the four basic operators, and parentheses. Operator precedence applies and division truncates toward zero.
Implement a Union-Find solution to find all connected components in an undirected graph. Return each component as a sorted list of node IDs.

“This one took me a minute to even figure out where to start.”

View Post

Snap·Machine Learning Engineer·Technical Phone Screen

Jul 2026

Snapchat ML Engineer interview with a coding round that was more of a math-precision trap than a pure DP problem. The question looked like standard coin change until you realized floating-point was the whole point.

Given a list of coin denominations as decimal values and a decimal target amount, find the minimum number of coins that sum exactly to the target. Each denomination can be used any number of times. Return -1 if no exact combination exists. You must avoid running DP directly on floats.

“My first instinct was to just throw DP at it with floats and call it a day.”

View Post

Snap·Machine Learning Engineer·Onsite - Behavioral / Leadership

Jul 2026

Behavioral round for an ML Engineer role at Snapchat, basically one big question about a hard project you led. Pretty standard format but the depth they expected was not.

Tell me about a challenging project you led or contributed to. What made it hard, what trade-offs did you navigate, how did you work with stakeholders, and what was the result? What would you do differently?

“This question is deceptively wide.”

View Post

Snap·Machine Learning Engineer·Onsite - System Design / Architecture

Jul 2026

Snap ML engineer interview with a system design question around building a healthcare agent for medical history summarization. Pretty open-ended, which I wasn't fully prepared for.

How would you design a system for a healthcare agent that summarizes a patient's medical history?

“I went straight to the model layer and spent too long talking about fine-tuning approaches before realizing they probably wanted to hear about data pipelines, privacy constraints, and how you'd handle unstructured clinical notes.”

View Post

Snap·Machine Learning Engineer·Technical Phone Screen

Jul 2026

Snapchat ML engineer loop, heavy on fundamentals. Five questions and every single one was a proper deep-dive, no warmup fluff. Left feeling like I'd been wrung out.

What is the relationship between cross-entropy and KL divergence, and can you derive why cross-entropy equals entropy plus KL divergence?
How does dropout create a mismatch between training and inference, and what are the practical ways to fix it? When would you use inverted dropout versus Monte Carlo dropout?
Why do large language models use LayerNorm rather than BatchNorm? What are the implications for variable sequence lengths, micro-batching, and training stability?
Compare SGD, SGD with momentum, and Adam: walk through the update rules, convergence properties, generalization behavior, and how you decide which to use.
Why does PPO use clipping or a KL penalty to constrain policy updates, and how does that stabilize training? What hyperparameters matter most and what are the common failure modes?

“I knew the punchline but fumbled the derivation live.”

View Post

Snap·Machine Learning Engineer·Technical Phone Screen

Jun 2026

Snapchat ML Engineer interview with a coding round that leaned more algorithm-heavy than I expected. The main problem was a classic DP string decoding question but they kept pushing with follow-ups until it got pretty uncomfortable.

Given a string of digits representing an encoded message where each number maps to a letter (1 to 26), count the number of distinct ways to decode it. Then optimize your solution to run in linear time with constant extra space. Follow-ups included returning the count modulo 1,000,000,007 for very large inputs, and extending the solution to handle streaming input where digits arrive one at a time.

“I got the basic DP solution out fast enough, two variables tracking the last two states, felt good.”

View Post

Snap·Machine Learning Engineer·Technical Phone Screen

Jun 2026

Snapchat ML engineer interview with a coding round that was pretty much a BFS grid problem. Nothing wild, but it's the kind of question that feels easy until you're actually writing it under pressure.

Given an n by n binary matrix, find the length of the shortest path from the top-left cell to the bottom-right cell, where you can move in all 8 directions and only through cells with value 0. Return -1 if no path exists.

“Classic BFS, and I knew it was BFS the moment I read it.”

View Post

Snap·Machine Learning Engineer·Technical Phone Screen

Jun 2026

Snapchat ML engineer interview with a classic puzzle-style coding problem that had a lot more depth than it looked. The follow-up questions pushed into edge cases and complexity analysis pretty hard.

You're given four numbers and a target value. Using addition, subtraction, multiplication, and division (with parentheses), and using each number exactly once, write a function that returns true if you can form an expression equal to the target. Follow-up: return a valid expression string if one exists. Also discuss division by zero handling, floating-point tolerance, your algorithmic approach, complexity, and how you'd extend this to N numbers or a different set of operators.

“I knew the backtracking approach going in but fumbled explaining the pruning part clearly.”

View Post