Uploaded September 2026 | Updated November 2026, 4 days ago
Nathan Drezner walks through deploying a Managed Deep Agent to production with Slack as the interaction layer. Starting from a research assistant agent already running in LangSmith Studio, he initializes a Slack channel, deploys, authorizes the connection, and shows how to monitor live traces before customizing the agent's name, icon, and color for the team.
Chapters:
0:00 What Managed Deep Agents are and why Slack matters
0:07 Meet the research assistant agent built in LangSmith Studio
0:18 Two ways to connect Slack: on creation or to an existing project
0:31 What slack init actually creates
0:45 Deploying for the first time
0:58 The Slack authorization step you'll hit on first deploy
1:23 Deployed! The agent sends its first message
1:41 Watching the request land in the LangSmith Studio dashboard
2:01 Monitoring traces when many users message your agent
2:19 Customizing your agent's name, icon, and color
3:09 Message triggers: manual tagging vs trigger on all messages
3:49 Trigger on all messages, live, and when to use it
4:46 More ways to run your Slack connected agent
Resources:
→ LangGraph: langchain.com/langgraph
→ LangSmith: langchain.com/langsmith
→ LangChain Academy: academy.langchain.com
Nathan Drezner walks through deploying a Managed Deep Agent to production with Slack as the interaction layer. Starting from a research assistant agent already running in LangSmith Studio, he initializes a Slack channel, deploys, authorizes the connection, and shows how to monitor live traces before customizing the agent's name, icon, and color for the team.
Chapters:
0:00 What Managed Deep Agents are and why Slack matters
0:07 Meet the research assistant agent built in LangSmith Studio
0:18 Two ways to connect Slack: on creation or to an existing project
0:31 What slack init actually creates
0:45 Deploying for the first time
0:58 The Slack authorization step you'll hit on first deploy
1:23 Deployed! The agent sends its first message
1:41 Watching the request land in the LangSmith Studio dashboard
2:01 Monitoring traces when many users message your agent
2:19 Customizing your agent's name, icon, and color
3:09 Message triggers: manual tagging vs trigger on all messages
3:49 Trigger on all messages, live, and when to use it
4:46 More ways to run your Slack connected agent
Resources:
→ LangGraph: langchain.com/langgraph
→ LangSmith: langchain.com/langsmith
→ LangChain Academy: academy.langchain.com










