Uber Machine Learning Engineer interview questions

33 Machine Learning Engineers have shared notes from their Uber interviews, with the questions they were asked in each round. That is 13% of the 263 Uber interview notes on Screna AI.

33 Machine Learning Engineer interview notes

Uber·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Uber MLE interview with a graph/BFS problem that looks deceptively straightforward until you actually try to implement it efficiently. Not the worst experience but definitely left me second-guessing my approach the whole way through.
  • Given an array of bus routes where each route loops forever, find the minimum number of buses you need to take to get from a source stop to a target stop. Return -1 if it's impossible.

“My first instinct was plain BFS on stops, which is wrong because the search space explodes.”

View Post

Uber·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Uber ML engineer screen, one coding question the whole time. It was a text generation problem that looked simple on the surface but had enough edge cases to trip you up if you weren't careful about how you built the frequency map.
  • Build a two-part text generation system: first a function that builds a word transition frequency map from a corpus, then a function that generates text by greedily picking the highest-frequency next word at each step.

“The frequency map part was fine, tokenize on whitespace, strip punctuation, nested dict.”

View Post

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

Jun 2026
ML system design round at Uber for an MLE role. The whole session was basically one big open-ended fraud detection design question that kept branching into deeper sub-topics. Pretty intense but not unfair.
  • Design an end-to-end machine learning system for fraud detection on payments or accounts.
  • Walk through the feature sources you'd use for a fraud detection model and explain the tradeoffs involved.
  • Compare gradient-boosted trees, random forests, and neural networks for fraud detection. What are the tradeoffs?
  • How would you handle the class imbalance problem in a fraud detection setting?
  • How should you structure your train/test split for a fraud model, and what pitfalls exist with standard random splits?
  • What metrics would you use to evaluate a fraud detection model, and why?
  • How would you scale training and serving for a fraud detection system when data volume exceeds what a single machine can handle?
  • How does your fraud model connect to user-facing actions, and how do you close the feedback loop back into training?

“This is the kind of question where you can spend 45 minutes and still feel like you barely scratched the surface.”

View Post

Uber·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Uber ML Engineer interview that was basically a deep dive into linear regression, not just the code but all the surrounding decisions. More thorough than I expected for a single problem.
  • Implement linear regression from scratch, including both a closed-form solution using the normal equation and a gradient descent solution with configurable learning rate and iterations. Also implement a predict method.
  • When would you prefer the closed-form normal equation over gradient descent, and what are the practical concerns around an ill-conditioned XᵀX matrix?
  • How would you add L2 regularization to both the closed-form and gradient descent versions of linear regression?
  • How would you evaluate the model fit using MSE and R-squared, and how would you distinguish between underfitting and overfitting?

“The closed-form part was fine, augmenting X with a ones column and inverting XᵀX is pretty mechanical.”

View Post

Uber·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Uber MLE phone screen, pretty much one coding problem the whole time with some follow-up discussion on scalability. Not a bad experience but the complexity conversation at the end was where I felt least prepared.
  • Given a string of digits (2-9), return all possible letter combinations using the standard phone keypad mapping. Then discuss how the output size scales for longer inputs and how you'd handle streaming the results.

“Got the backtracking solution down fine, DFS with a running path and appending to results at the leaf nodes.”

View Post

Uber·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Uber ML engineer round that was pretty much one focused coding question on implementing CLIP-style contrastive loss. Felt like a niche topic if you haven't touched multimodal stuff before, but the math itself isn't too bad once you see the pattern.
  • Implement a contrastive loss function for a vision-language model given a batch of image and text embeddings. The loss should be symmetric: compute cross-entropy over the similarity matrix for images, do the same transposed for texts, then average them.

“I knew the general idea behind contrastive learning but had never written this specific formulation from scratch under pressure.”

View Post

Uber·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Uber ML engineer interview, technical phone screen focused entirely on implementing logistic regression from scratch. Pretty deep dive, they wanted working code plus a discussion of the math and engineering tradeoffs. Felt more like a mini take-home crammed into a live session.
  • Implement logistic regression from scratch in Python. It needs to actually run end-to-end, including fit, predict_proba, and predict methods using gradient descent on the negative log-likelihood loss.
  • How do you choose a learning rate, decide when the model has converged, and incorporate L2 regularization into logistic regression?

“The 'must run end-to-end' part is what tripped me up initially.”

View Post

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

Jun 2026
Uber ML Engineer onsite, system design round focused entirely on feed ranking. The scope was broad enough that I kept second-guessing whether to go deeper on one stage or keep moving. Left feeling okay but not confident.
  • Design a personalized feed ranking system that returns an ordered list of items from a large content corpus, accounting for user history, social graph, freshness, and multiple competing objectives like engagement, diversity, and content integrity.

“This is one of those questions that looks scoped until you realize the surface area is enormous.”

View Post

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

Jun 2026
ML system design round at Uber for an MLE role. The whole session was a single deep-dive into building a destination autocomplete system, and it went pretty far into the weeds on ranking, latency, and personalization tradeoffs.
  • Design an ML system that predicts and ranks destination suggestions in the Uber app as a user types the first few characters into the search bar.
  • What features would you use in the ranking model, and would you go with a gradient-boosted model or a neural ranker?
  • How would you handle cold-start for new users who have no trip history?
  • What offline and online metrics would you use to evaluate this system, and how do they relate to each other?
  • How do you think about the privacy and personalization tradeoff in a system like this?

“This is the whole question, basically.”

View Post

Uber·Machine Learning Engineer·Technical Phone Screen

Jun 2026
ML engineer round at Uber focused on the Eats homepage recommendation system. The interviewer had an economics/operations research background and didn't actually work on recommendations, which made the whole thing feel off. Pretty sure I didn't pass.
  • How would you design a restaurant recommendation system for the Uber Eats homepage?

“The mismatch between what I prepared for and who was actually interviewing me was rough.”

View Post