AI Engineering Interview Questions — Practice Library

42 problems
Problem
Tags
Type
GPU Inference Batching System Design with Synchronous User Requests
Reported at AnthropicML EngineeringHard
System Design
Distributed Search System Architecture with LLM Inference at Scale
Reported at AnthropicML EngineeringHard
System Design
Load balancing for distributed AI model serving and inference requests
ML EngineeringMedium
Theory
Medical LLM Governance: Audit Infrastructure at Meta Scale
Reported at MetaML EngineeringHard
Theory
Model Benchmarking: Metrics, Traffic Split, and Rollout Strategy
Reported at NetflixML EngineeringMedium
Theory
Recommendation Model Deployment — Canary Failure
ML EngineeringMedium
Theory
End-to-end LLM query batching system design and optimization
Reported at AnthropicML EngineeringHard
System Design
Elbow Method and Silhouette Score for Optimal Cluster Selection
Reported at MicrosoftML EngineeringEasy
Theory
Model Evaluation for Imbalanced Loan Fraud Detection
ML EngineeringEasy
Theory
Detect Fraudulent Reviews with Machine Learning Classification
Reported at MicrosoftML EngineeringMedium
System Design
ML Experiment Tracking System with Metrics Analysis
Reported at Google DeepMindML EngineeringHard
System Design
Feature Store Architecture for Real-Time Fraud Detection
ML EngineeringEasy
System Design
Average Product Rating Calculation and Top Five Identification
Reported at LinkedInML EngineeringEasy
Coding
Distributed Training Pipeline for Trillion-Parameter Language Model
Reported at NvidiaML EngineeringHard
System Design
Sketch a complete LLMOps pipeline from raw data to serving to feedback
ML EngineeringHard
System Design
Housing Price Outlier Detection and Visualization System
Reported at LinkedInML EngineeringEasy
Coding
Data quality strategies for large language model training pipelines
Reported at NvidiaML EngineeringMedium
System Design
How do you monitor performance drift and hallucinations in production LLMs?
ML EngineeringHard
System Design
Handling class imbalance: techniques for fair model training
Reported at JP Morgan ChaseML EngineeringEasy
Theory
Scalable Token Generation Service Architecture for High-Throughput LLM
Reported at AnthropicML EngineeringHard
System Design
CI/CD for LLM workflows — what is different from traditional ML?
ML EngineeringHard
System Design
Data Cleaning Strategies for Handling Messy and Incomplete Datasets
Reported at AmazonML EngineeringMedium
Theory
Handling Class Imbalance and Preventing Overfitting in Machine Learning
Reported at AmazonML EngineeringMedium
Theory
Model Monitoring — Diagnosing Drift in a Credit Scoring Model
ML EngineeringHard
Theory
Loading more problems…