The long-form reference.
Every chapter of the open-source AI System Design Guide, from foundations to case studies and interview prep. Read the full text here, then jump to the animated lesson for any concept.
Written by Om Bharatiya and contributors, MIT licensed. Included here with attribution; each chapter links back to its source.
Interview Prep
Question bank, answer frameworks, whiteboard exercises and the 2026 job market.
- Interview Prep — Overview5m
- AI System Design Interview Question Bank443m
- Answer Frameworks for AI System Design Interviews6m
- Common Pitfalls in AI System Design Interviews10m
- Whiteboard Exercises for AI System Design14m
- Behavioral Questions for AI/ML Roles20m
- AI Job Market Trends - November 202652m
- Frequently Asked Questions: AI Engineering, RAG, and Agents28m
Foundations
LLM internals, tokenization, attention, transformers, embeddings, inference.
Model Landscape
Model taxonomy, capability assessment, pricing and model selection.
Training & Adaptation
Pretraining, fine-tuning, PEFT, RLHF/DPO, distillation, quantization, RLVR.
Inference Optimization
KV cache, speculative decoding, batching, paged attention, serving, cost.
Prompting & Context
Prompt engineering, ICL, CoT/ToT, context engineering, structured output, DSPy, injection defense.
Retrieval Systems
RAG end to end: chunking, embeddings, vector DBs, hybrid search, reranking, GraphRAG and more.
- RAG Fundamentals8m
- Chunking Strategies5m
- Embedding Models9m
- Vector Databases20m
- Hybrid Search8m
- Reranking Strategies11m
- GraphRAG8m
- Agentic RAG7m
- Advanced Retrieval Patterns4m
- Contextual Retrieval13m
- Late Interaction & ColBERT11m
- Multi-Modal RAG14m
- RAG Evaluation Patterns18m
- Production RAG at Scale20m
- Data Engineering for AI14m
Agentic Systems
Agent fundamentals, reasoning loops, MCP, multi-agent, planning, HITL, durable execution.
- Agentic Systems — Overview3m
- Agent Fundamentals5m
- Reasoning Loops: ReAct and Beyond3m
- Tool Use and MCP46m
- Multi-Agent Orchestration10m
- Agent Memory and State26m
- Planning and Decomposition5m
- Error Handling and Recovery5m
- Human-in-the-Loop Patterns12m
- Agentic Security and Sandboxing23m
- Evaluating Agentic Systems10m
- Durable Execution for Long-Running Agents13m
- Loop Engineering22m
Memory & State
Memory architectures, short/long-term memory, semantic caching, state management.
Frameworks & Tools
LangChain, LangGraph, LlamaIndex, DSPy, Semantic Kernel, AutoGen/CrewAI, coding agents.
- LangChain Deep Dive7m
- LangGraph Orchestration5m
- LangSmith Observability6m
- LlamaIndex6m
- DSPy: Programming Language Models6m
- Semantic Kernel3m
- Microsoft Agent Framework, CrewAI, and the Agent SDK Landscape17m
- Framework Selection Guide11m
- Claude Code: The Autonomous Coding Agent19m
- OpenCoder: AI Coding Agents Landscape24m
- Pydantic AI and Mastra: Typed Agent Frameworks (2026)12m
- Navigating Framework Churn21m
Document Processing
OCR and layout analysis.
Infrastructure & MLOps
LLM infrastructure, CI/CD, AI gateways, FinOps.
Security & Access
LLM security and access control.
Reliability & Safety
Guardrails, ensembles, reliability patterns, governance.
Evaluation & Observability
LLM evaluation, observability, benchmarks.
AI Design Patterns
Design patterns and anti-patterns.
Case Studies
Twenty production systems, designed end to end.
- Case Study: Enterprise RAG System8m
- Case Study: Customer Support Conversational Agent7m
- Case Study: Financial Analysis with Ensemble Verification5m
- Case Study: AI Code Assistant7m
- Case Study: Content Moderation at Scale6m
- Case Study: Real-Time AI Search Engine4m
- Case Study: Autonomous Coding Agent10m
- Case Study: Multi-Tenant AI SaaS Platform7m
- Case Study: AI-Powered Customer Support5m
- Case Study: Document Intelligence Pipeline5m
- Case Study: AI-Powered Recommendation Engine4m
- Case Study: Regulatory Compliance Automation3m
- Case Study: Voice AI Assistant for Healthcare7m
- Case Study: Real-Time Fraud Detection3m
- Case Study: Enterprise Knowledge Management3m
- Case Study: Production Computer-Use Agent15m
- Case Study: Multi-Tenant Fine-Tuning Platform14m
- Case Study: Eval-Gated CI/CD for an AI Product13m
- Case Study: Customer-Specific Distillation Pipeline12m
- Case Study: Enterprise MCP Knowledge Agent16m
Tool Use & Computer Agents
Tool-use landscape, architectures, computer-use agents, safety.
Voice & Audio Agents
Real-time voice agents.
Multimodal Generation
Image, video and audio generation systems.
References & Evals
Glossary, patterns catalog, research radar, transition guide and two complete evals study guides.