langgraph and microsoft azure durable functions

Question

I’m currently exploring options for building AI agents and have come across both LangGraph and Microsoft Azure Durable Functions. It seems that both can orchestrate workflows, but I’m trying to understand their differences and respective roles in this context.

Specifically, I’m interested in:

1. How does LangGraph handle workflow orchestration for AI agents compared to Azure Durable Functions?
2. What are the key differences in their architectures and capabilities? In what scenarios would one be preferred over the other for building stateful, scalable AI applications?

Any insights or experiences with these technologies would be greatly appreciated!