We’re releasing a new LangChain  Academy course, LangGraph Essentials,  

where you can learn the basics of  LangGraph in less than an hour.

LangGraph is a low-level orchestration framework  designed specifically for building AI agents. It  

provides a durable runtime for agents with  graph-based execution. LangGraph allows you  

to create flexible, agentic workflows with  its modular components. It allows you to  

control execution, manage state, allow for human  intervention when needed, and scale reliably.

LangGraph addresses the unique challenges  of building with large language models. It  

addresses latency by offering streaming and  parallelism. It addresses reliability with  

checkpointing, which allows for retries  and human intervention when needed.

LangGraph builds upon feedback we’ve received from  the broad adoption of the LangChain framework,  

and rethinks how agent frameworks should work  for production. We wanted to find the right  

abstraction for AI agents, and decided that it  was little to no abstraction at all. Instead,  

we focused on control and durability – with  the goal of making LangGraph what you’d use  

to run your agents in production. We’ve now  seen this work with companies like Linkedin,  

Uber, and Klarna using LangGraph to  power their agents in production.

In this quickstart course, you’ll explore  the essential building blocks of LangGraph  

by creating simple workflows, and then  by building a sample agent to tie it all  

together. I hope that you can take the time  to learn more about LangGraph in this course.