All videos
0:00 / 0:00
event

Interrupt 26: The Agent Conference by LangChain

LangChain21 July 2026Watch on YouTube

Part of series

Ep. 9 · LangChain Deep Agents

Stap-voor-stap uitleg en tutorials over het bouwen, configureren en deployen van Deep Agents met het LangChain-framework.

View the series

What you'll learn

  • The central theme of Interrupt 2026 is the move from agent prototypes to production agents that require their own infrastructure.
  • LangChain launches two products at the conference at once, the SmithDB database and the LangSmith Engine improvement agent.
  • Trace payloads are growing fast, from 6KB to 37KB at P50 and from 364KB to 12MB at P99 over two years.
  • According to LangChain, performance matters not only for humans but also for the agents themselves, which slow down from inefficient tools.
  • The event draws a broad production audience with speakers from Toyota, LinkedIn, Coinbase and Etsy among others.

Frequently asked questions

What is Interrupt 2026?
Interrupt is LangChain's flagship agent conference, held on May 13-14, 2026 in San Francisco. The event revolves around bringing agents into production and includes keynotes, product launches and production stories from companies such as Toyota and Coinbase.
Which products is LangChain launching at Interrupt 2026?
LangChain introduces SmithDB, a database powering the core LangSmith workloads, and LangSmith Engine, an agent that clusters failures and suggests improvements automatically. Both are available in public beta.
Why do trace performance levels matter for agents?
According to LangChain co-founder Harrison Chase, performance matters not only for human UX but also for the agents themselves. Slow tools and observability hold back the agents that depend on them, which affects the entire production environment.

Topics

ai agentslangchaindeveloper toolsproduction deploymentconference

Sources

What is known about this topic outside the broadcast, and where it says so.

Description from the channel

Interrupt is the conference for engineering and AI leaders, software engineers, ML engineers, and researchers building and deploying agents. Whether you're leading AI initiatives or hands-on with the code, you'll find content designed for your needs, from strategic insights to technical deep dives that will equip you to build and manage agents in production. Learn more about the agent conference by LangChain: https://interrupt.langchain.com/ Chapters: 0:00 A year ago, building agents was hard 0:25 LangChain's plan for the agent development lifecycle 0:37 Why agent traces are the center of agent development 1:14 Launching SmithDB and LangSmith Engine 1:32 More than a platform