19 Core Concepts of AI System Design: From KV Cache to Multi-Agent Systems

blaizedsouza · x · 2026-11-19

A curated map of 19 concept clusters for AI system design: LLM fundamentals and inference servers, AI hardware (GPU/TPU/LPU), tokens, latency and throughput, prefill/decode/KV cache, scaling and load balancing, caching and LLM routing, embeddings and vector databases, RAG, context engineering, streaming and AI gateways, agents with tool calling/MCP/memory, multi-agent and A2A, multimodal and voice AI, safety guardrails, evaluation and observability, cost optimization, fine-tuning and inference optimization, plus AI system design interviews.

Original post →

More from coding & agent

coding & agent channel →