Free
ML & AI·Personalization
Design a recommendation system
Recommend useful content while balancing relevance, freshness, and exploration.
Advanced 60 min
ML system design interview questions
ML system design interviews go beyond the model. These rounds ask you to design recommendation systems, feature stores, distributed training platforms, low-latency model serving, real-time fraud scoring and retrieval-augmented generation.
Practice the parts interviewers probe: training and serving skew, feature freshness, offline and online evaluation, latency budgets, and how the system degrades when a model or data source fails.
Plus
ML & AI·ML platforms
Design a feature store
Serve consistent ML features for both training jobs and online predictions.
Advanced 60 min
Plus
ML & AI·Training systems
Design a distributed model training platform
Schedule reproducible training jobs across shared compute and data.
Advanced 60 min
Plus
ML & AI·Inference
Design a low-latency ML model serving platform
Deploy models safely and handle variable inference traffic.
Advanced 60 min
Plus
ML & AI·Generative AI
Design a retrieval-augmented generation system
Answer questions from a changing document collection with traceable sources.
Advanced 60 min
Plus
ML & AI·Fraud & risk
Design real-time fraud scoring
Score transactions in milliseconds and route uncertain cases for review.
Advanced 60 min
Plus
ML & AI·Model quality
Design an ML model monitoring system
Notice when a production model becomes slow, biased, or less useful.
Intermediate 40 min