Building Effective Agents with LangGraph

Anthropic's recent blog post on "Building Effective Agents" lays out the difference between "agents" and "workflows", and presents a number of common patterns for both. Here, we implement every workflow and agent pattern covered in the blog from scratch using LangGraph. We explain the key differences between workflows and agents, when to use each approach, and how to implement them effectively. We also cover the benefits you can gain from using LangGraph as a framework. Documentation: https://langchain-ai.github.io/langgraph/tutorials/workflows/ Video notes: https://mirror-feeling-d80.notion.site/Workflow-And-Agents-17e808527b1780d792a0d934ce62bee6?pvs=4 Timestamps: 0:00 Introduction & Key Concepts 1:00 Understanding Workflows vs Agents 2:00 Why Use Frameworks? Benefits of LangGraph 4:00 Building Block: Augmented LLM 5:00 Pattern 1: Basic Prompt Chaining 9:00 Pattern 2: Parallelization 11:00 Pattern 3: Routing with LLMs 14:00 Pattern 4: Orchestrator-Worker Pattern 20:00 Pattern 5: Evaluator-Optimizer Workflow 24:00 Building Agents: Beyond Workflows 27:00 Implementing a Basic Agent Loop 30:00 Conclusion & LangGraph Benefits