Uber
Ride-hailing and delivery platform with massive real-time data challenges.
4 Rounds
~21 Days
Hard
The Interview Loop
Recruiter Screen (30 min)
Standard fit check, behavioral questions, and resume overview.
Technical Loop (3-4 Rounds)
Deep dive into domain knowledge, coding, and system design.
Interview Question Bank
Machine Learning Engineer
•
Behavioral
•
medium
Tell me about a time a machine learning model you deployed degraded in production or caused a business metric to drop. How did you detect it, and what steps did you take to resolve it?
#Incident Response
#Ownership
#Monitoring
Machine Learning Engineer
•
Behavioral
•
medium
Describe a situation where you believed a complex ML model was necessary, but the engineering or product team wanted a simple heuristic. How did you handle the disagreement?
#Communication
#Trade-offs
#Collaboration
Machine Learning Engineer
•
Coding
•
medium
Given an array of ride requests with start and end times, find the minimum number of drivers required to fulfill all requests.
#Arrays
#Sorting
#Greedy
#Heap
Machine Learning Engineer
•
Coding
•
hard
Given a city map represented as a weighted directed graph where nodes are intersections and edges are roads with travel times, write a function to find the top K shortest paths between a rider and a destination.
#Graphs
#Dijkstra
#A* Search
#Priority Queue
Machine Learning Engineer
•
Coding
•
medium
Implement a rate limiter for the Uber API to prevent abuse. The rate limiter should allow a maximum of N requests per user per minute.
#Data Structures
#Concurrency
#Sliding Window
Machine Learning Engineer
•
Coding
•
medium
Implement an autocomplete system for the Uber app destination search. Given a prefix, return the top 5 most frequently searched locations.
#Trie
#Hash Map
#String Manipulation
#Design
Machine Learning Engineer
•
System Design
•
hard
Design Uber's surge pricing model. How would you predict demand and supply in a given geohash for the next 10 minutes?
#Time Series
#Geospatial Data
#Regression
#Real-time ML
Machine Learning Engineer
•
System Design
•
hard
Design a restaurant recommendation system for Uber Eats. How do you handle the cold start problem for new users and new restaurants?
#Recommendation Systems
#Collaborative Filtering
#Two-Tower Models
#Cold Start
Machine Learning Engineer
•
System Design
•
medium
Design a real-time fraud detection system for Uber accounts (e.g., detecting stolen credit cards or fake driver accounts). How do you deal with highly imbalanced data?
#Anomaly Detection
#Classification
#Imbalanced Data
#Graph Neural Networks
Machine Learning Engineer
•
System Design
•
hard
Design the ML architecture for dispatching drivers to riders. How do you balance minimizing rider wait time with maximizing driver earnings and system throughput?
#Reinforcement Learning
#Bipartite Matching
#Operations Research
#Multi-objective Optimization
Machine Learning Engineer
•
Technical
•
medium
Uber uses deep learning for ETA prediction. If you notice the ETA model is consistently under-predicting travel times during heavy rain, how would you debug and fix this?
#Model Debugging
#Feature Engineering
#Data Drift
#Deep Learning
Machine Learning Engineer
•
Technical
•
hard
We want to roll out a new matching algorithm that pairs drivers and riders. How would you design the A/B test? What metrics would you track, and how do you handle network effects (interference) between the control and treatment groups?
#A/B Testing
#Network Effects
#Switchback Testing
#Statistics
Machine Learning Engineer
•
Technical
•
medium
For predicting the exact location (latitude/longitude) of a rider's pickup spot based on historical pin drops, what loss function would you use and why? How would you handle outliers?
#Loss Functions
#Geospatial
#Regression
#Robust Statistics
Machine Learning Engineer
•
Technical
•
medium
How would you generate embeddings for Uber Eats restaurants to capture semantic similarities (e.g., a user who likes 'Spicy Thai' might like 'Sichuan Chinese')? Describe the data and architecture.
#Embeddings
#NLP
#Word2Vec
#GraphSAGE
Machine Learning Engineer
•
Technical
•
easy
Uber heavily uses XGBoost for tabular data but Deep Learning for ETA and NLP. Explain the theoretical differences between Gradient Boosted Trees and Neural Networks. When would you strictly prefer one over the other?
#XGBoost
#Deep Learning
#Model Selection
#Bias-Variance Tradeoff
Difficulty Radar
Based on recent AI-sourced data.
Meet Your Interviewers
The "Standard" Interviewer
Senior EngineerFocuses on core competencies, system constraints, and clear communication.
SimulateUnwritten Rules
Think Out Loud
Always explain your thought process before writing code or drawing architecture.