System Design - Netflix & Kimi - Building Large Scale AI & Distributed Systems

How do systems like Netflix serve millions of users at massive scale? And how does an AI system like Kimi handle large-scale model inference? In this session, we break down the system design and architecture of Netflix and Kimi, looking at the engineering principles behind large-scale distributed systems and AI infrastructure. We cover: • Netflix system architecture and large-scale video streaming • Content delivery, caching, storage and databases • Load balancing and distributed services • Scalability, availability and fault tolerance • Kimi's AI/LLM system architecture • Large-scale inference and GPU infrastructure • Distributed inference and scheduling • KV cache and memory considerations • How modern AI systems differ from traditional distributed systems • Key system-design trade-offs and architectural decisions The goal is to understand not just what components exist, but why these architectures are designed the way they are and how they scale. This session connects traditional distributed-system design with modern AI infrastructure and large-scale LLM serving. #SystemDesign #Netflix #Kimi #DistributedSystems #AIInfrastructure #LLM #SystemArchitecture #MachineLearning #Inference #GPU