0:00 / 0:00
ai
Managed Deep Agents - Quickstart
LangChain19 August 2026Watch on YouTube
Part of series
Ep. 7 · Actually Agent Agents
View the seriesWhat you'll learn
- Learn how to set up a Managed Deep Agent using LangChain, from project scaffolding to full deployment.
- Model configuration and instructions are essential for your agent's behavior and are explained step by step.
- You can add search functionality to your agent for better information retrieval and more relevant answers.
- LangSmith Studio provides a testing environment where you can validate your agent before deploying it.
- The MDA CLI enables simple deployment to a hosted runtime without complex infrastructure configuration.
Frequently asked questions
What are Managed Deep Agents and what do they support?
Managed Deep Agents are AI agents that allow you to automate tasks via LangChain. They provide their own harness and hosted runtime, so you don't have to set up complex infrastructure yourself.
What steps do you follow to set up a Managed Deep Agent?
You scaffold the project, configure the model and instructions, add search functionality, test everything in LangSmith Studio, and then deploy with the MDA CLI to the hosted runtime.
Why is LangSmith Studio important in this process?
LangSmith Studio provides a testing environment where you can validate your agent and verify everything works correctly before deploying it live.
How do you deploy a Managed Deep Agent after testing it?
You use the MDA CLI command to deploy your agent to the hosted runtime without setting up your own servers or infrastructure.
Topics
Description from the channel
Create and deploy your first Managed Deep Agent: scaffold a project, configure the model and instructions, add search, test in LangSmith Studio, and deploy with the MDA CLI. Managed Deep Agents supplies the Deep Agents harness and hosted runtime. Docs: https://docs.langchain.com/langsmith/python/managed-deep-agents-quickstart LangSmith: smith.langchain.com