Overview
RemoteGraph allows you to call LangGraph graphs deployed on LangSmith directly from emma. This enables you to leverage your custom AI workflows and agents within emma’s automation capabilities.What is RemoteGraph?
RemoteGraph is a LangGraph feature that allows you to interact with graphs deployed on LangSmith as if they were running locally. By integrating RemoteGraph with emma, you can:- Execute custom AI workflows - Run your LangGraph agents from emma
- Leverage LangSmith deployments - Use graphs deployed on LangSmith infrastructure
- Combine with Automations - Schedule and automate your LangGraph workflows
Getting Started
1
Deploy your graph to LangSmith
First, deploy your LangGraph graph to LangSmith. Refer to the LangSmith documentation for deployment instructions.
2
Get your deployment URL
After deployment, obtain your LangSmith deployment URL and API key.
3
Configure in emma
Add your RemoteGraph configuration in emma’s settings.
4
Start using
Call your RemoteGraph from emma tasks or automations.
Configuration
To connect your LangSmith deployment to emma, you’ll need:Use Cases
Custom AI Agents
Run your custom LangGraph agents through emma
Complex Workflows
Execute multi-step AI workflows with state management
Scheduled Execution
Combine with Automations for scheduled graph execution
Team Collaboration
Share deployed graphs across your team via emma
Best Practices
Monitor your deployments
Monitor your deployments
Use LangSmith’s monitoring features to track graph execution and performance.
Handle errors gracefully
Handle errors gracefully
Implement proper error handling in your graphs for reliable automation.
Optimize for latency
Optimize for latency
Consider response times when using RemoteGraph in time-sensitive automations.