LangGraph Foundations: Graph-Based LLM Prompting and Chaining Training Course

Introduction to LangGraph and Graph Theories

• The rationale for using graphs in LLM apps: orchestration versus simple chains
• The role of nodes, edges, and state in LangGraph
• Getting started: building the first executable graph

State Management and Prompt Sequencing

• Structuring prompts as distinct graph nodes
• Transferring state between nodes and managing outputs
• Memory strategies: distinguishing short-term vs. persistent context

Branching Logic, Control Flow, and Error Resolution

• Implementing conditional routing and multi-path processes
• Managing retries, timeouts, and fallback mechanisms
• Ensuring idempotency and safe re-execution

Tool Utilization and External Connectivity

• Invoking functions and tools from within graph nodes
• Interacting with REST APIs and services inside the graph
• Handling structured outputs effectively

Retrieval-Augmented Processes

• Basics of document ingestion and chunking
• Utilizing embeddings and vector stores (e.g., ChromaDB)
• Generating grounded answers with citations

Testing, Troubleshooting, and Assessment

• Conducting unit-style tests for nodes and pathways
• Implementing tracing and observability measures
• Performing quality checks: factuality, safety, and consistency

Essentials of Packaging and Deployment

• Setting up environments and managing dependencies
• Serving graphs via APIs
• Versioning workflows and managing rolling updates

Conclusion and Future Directions