The LangGraph Platform consists of several components that work together to support the development, deployment, debugging, and monitoring of LangGraph applications. LangGraph Studio is a central components of this and is a specialized IDE to enable visualization, interaction, and debugging of LangGraph applications. Here, we explain how Studio works and review a few common usage patterns, such as threads, assistants, and debugging with LangSmith.
Docs:
https://langchain-ai.github.io/langgraph/concepts/langgraph_studio/
Video notes:
https://mirror-feeling-d80.notion.site/Tour-of-LangGraph-Studio-1bd808527b17807eb995da7fe824a281?pvs=4
Chapters:
00:00 - Introduction to LangGraph Studio
00:21 - Demo: Local Deep Researcher in Studio
00:56 - Understanding Graph Execution in Real-time
01:17 - LangGraph Platform Components Overview
02:07 - Memory in LangGraph Platform
02:35 - Studio's Role in the Platform
03:00 - Application Structure Requirements
03:25 - Configuration File Overview
04:05 - Application Structure Components
04:46 - Running LangGraph Platform Locally
05:19 - Using uvx for Environment Management
05:55 - Launching the Local Development Server
06:17 - Exploring API Documentation
06:42 - Studio UI Introduction
07:07 - Graph Visualization and Node Mapping
07:55 - Understanding Input and State Management
08:46 - Running a Graph in Studio
09:17 - State Updates During Execution
09:58 - Thread Management and History
10:46 - Memory Button and LangGraph Store
11:14 - LangSmith Integration
11:33 - Thread View in LangSmith
12:10 - Assistant Configurations Management
12:44 - Creating New Assistants
13:27 - Configuration Settings Customization
14:00 - Studio Features Recap
14:48 - Conclusion