LangChain - Videos
Back to Channel
Tracing Claude Code to LangSmith
Curious what Claude Code is doing behind the scenes? Or want observability into critical workflows that you’ve set up with Claude Code With our new Claude Code → LangSmith integration, you can vie...
Approaches for Managing Agent Memory
Humans refine their skills and learn preferences through experience. But many AI agents lack this capacity for continual learning. Here, we give an overview of memory in the DeepAgents CLI. Here, w...
LangChain Academy New Course: Introduction to LangChain - Python
Learn how to build with LangChain – our open source framework that makes it easy to start building agents with any model provider. In this course, you’ll create agents that can reason, use tools, ...
Build an MCP Agent with Claude: Dynamic Tool Discovery Across Cloudflare MCP Servers
In this video, Christian Bromann build and demo an agent that uses LLM provider native tools—specifically Anthropic Claude’s built-in MCP toolsets—to connect to Cloudflare’s managed MCP servers. I...
The agent development loop with LangSmith + Claude Code / Deepagents
LangSmith provides a system of record for traces, which can include long agent trajectories. Code agents like Claude Code or Deep Agents CLI can make use of traces to improve agent performance via ...
I Let an AI Control My Browser to Play Tic-Tac-Toe - LangChainJS Tutorials
What if an AI could **actually use the browser** — not through brittle scripts, but by *seeing* the UI and deciding where to click? In this video, I explain how modern **agent tools** work and dem...
Building & Observing a Deep Agent for Email Triage with LangSmith
In this video, we walk through how to build and observe a deep agent using LangSmith. We’ll build a simple email assistant that reads incoming emails and decides how to handle them — triage, respo...
Observing & Evaluating Deep Agents Webinar with LangChain
Explore the unique challenges of observing and evaluating Deep Agents in production. Deep Agents represent a shift in how AI systems operate – unlike simple chatbots or basic RAG applications, thes...
Trace OpenRouter Calls to LangSmith — No Code Changes Needed
OpenRouter's new Broadcast feature lets you send all your traces directly to LangSmith with no code changes required — whether you’re tracing with LangChain, provider SDKs, or the OpenRouter SDK. ...
LangSmith Fetch: CLI tool to debug agents from your terminal
Learn how to debug agents from your terminal using LangSmith Fetch, the new CLI that brings LangSmith trace data directly into your workflow. Pull traces and threads into your terminal or IDE with ...
Polly: The AI Assistant for AI Engineering in LangSmith
See how Polly, the AI assistant for AI engineering in LangSmith, helps you understand long-running agent executions by analyzing traces, threads, and prompts. We’ll walk through: • Debugging compl...
How to debug voice agents with LangSmith
Learn how to debug and improve a AI voice agent using LangSmith. We’ll walk through tracing conversations, spotting failures, and iterating on your agent. In this demo we use LangChain and Pipecat...
Build a voice agent with LangChain
Learn how to build a speech-to-text voice agent using LangChain. We break down the two methods of building voice agents and several of the key considerations for using each architecture: 1. STT / ...
Production-Ready Agents: Automatic Tool Retries with Exponential Backoff
Tools fail. APIs time out. Services throw random 500s. If your agent can’t recover, your entire workflow collapses. In this tutorial, Christian Bromann walks through how to use Tool Retry Middlewa...
Building a Linear issue agent with Langsmith Agent Builder
Learn how to build an agent that can create and edit Linear issues using our no-code Agent Builder. Try it for free today: https://langsmith.com/
Building a Market Research Assistant with Langsmith Agent Builder
Learn how to quickly create a research agent in LangSmith using our no-code Agent Builder. Try it for free today: https://langsmith.com/
Building an Email Assistant with Langsmith Agent Builder
Learn how to quickly create an email assistant in LangSmith using our no-code Agent Builder. Try it for free today: https://langsmith.com/
Summarization Middleware (Python)
Learn about how to use LangChain's summarization middleware as a key component of your context engineering pipeline. This middleware is automatically triggered and helps to keep your long running a...
LangSmith Agent Builder Now Available in Public Beta
LangSmith Agent Builder is now in Public Beta, enabling anyone to create production ready agents without writing code. Learn what's new in the Beta release, how people are using Agent Builder, how...
Anthropic-Style Context Editing… Now for Every LLM in LangChainJS!
Learn how to bring Anthropic’s powerful Context Editing capabilities — like tool result clearing — to any LLM using LangChainJS. In this video, Christian Bromann breaks down Anthropic’s original d...
Product Evals (for AI Applications) in Three Simple Steps
Eugene Yan wrote a GREAT blogpost on how to build product evals (for AI applications) in three simple steps: https://eugeneyan.com/writing/product-evals/ These three steps are: 1. Label data 2. Al...
AI Agents in Production: Lessons from Rippling and LangChain
How does a company deploy AI agents across HR, payroll, IT, and finance products used by thousands of companies? Ankur Bhatt, Head of AI at Rippling, shares insights on building production-ready ag...
Using skills with Deep Agents CLI
Anthropic recently introduced the idea of agent skills, a directory of folders that an agent can access to perform different actions. Here, we talk about skills, why they are interesting, how agent...
Managing Agent Context with LangChain: Summarization Middleware Explained
Long-running agents eventually hit context overload — leading to context poisoning, distraction, confusion, and degraded performance. In this video, Christian from LangChain breaks down how Summar...
What are Deep Agents?
Deep Agents is a term we coined to describe agents capable of handling complex, open-ended tasks over long time horizons. We identified four essential components that make this possible: a planning...
Build a Research Agent with Deep Agents
Deepagents is a simple, open source agent harness built by LangChain. It uses some common principle seen in popular agents such as Claude Code and Manus, including planning (prior to task execution...
Model Call Limit Middleware (Python)
Learn about how to use LangChain's model call limit middleware as a guardrail for agents. This middleware can be used for single agent invocations or across conversations. We walk through an exampl...
Agents Gone Wild? Use Tool Call Limits in LangChainJS to Keep Them in Check!
In this tutorial, Christian Bromann will show you how to prevent runaway tool usage using the Tool Call Limit Middleware in LangChainJS. You’ll learn how to set clear, declarative limits on tool us...
Building a Research Agent with Gemini 3 + Deep Agents
Google's eagerly anticipated Gemini 3 Pro release shows state-of-the-art performance across a wide range of agentic evaluations. Here, we show how to use Gemini 3 Pro with Deepagents, our open sou...
Model Fallback Middleware (Python)
Learn about how to use LangChain's new model fallback middleware to add resilience to your applications. Automatically switch between models and even providers when an API call fails with just a fe...
Stop Endless Back-and-Forth — Add Model Call Limits in LangChainJS
Is your support agent getting stuck in endless back-and-forth conversations? Users asking the same thing three different ways? Or long chats where the bot clearly isn’t helping anymore? In this tu...
LangChain Academy New Course: LangSmith Essentials
Testing applications is essential to the development lifecycle, but LLM systems are non-deterministic – you can’t always predict how they will behave. Add multi-turn interactions and tool-calling...
To-Do List Middleware (Python)
Learn about how to use LangChain's to-do list middleware to equip agents with task planning and tracking capabilities for complex multi-step tasks. Our example uses a software development agent. M...
Why Most AI Agents Fail — and How a Simple Todo List Fixes It
Most AI agents don’t fail because the model is bad — they fail because they don’t plan. In this video, @christian-bromann from LangChain shows how the TodoListMiddleware gives your agents structure...
Execute code with sandboxes for Deep Agents
We're excited to launch Sandboxes for DeepAgents, a new set of integrations that allow you to safely execute arbitrary code and bash commands in remote sandboxes. Your DeepAgent runs locally (or wh...
Add a Human-in-the-Loop to Your LangChain Agent (Next.js + TypeScript Tutorial)
Bring humans back into the loop 👩💻 — this tutorial shows how to integrate Human-in-the-Loop (HITL) middleware into your LangChainJS agents using createAgent. You’ll learn how to: - Pause agent ex...
Tool Call Limit Middleware (Python)
Learn about how to use LangChain's tool call limit middleware to control an agent's tool calling abilities across single interactions and full conversations. Our example uses a shopping agent that...
How Agents Use Context Engineering
This video covers the core principles of context engineering for AI agents and how they're implemented across popular frameworks like Claude Code, Manus, and LangChain's DeepAgents. As AI agents ta...
Building a Typescript deep research agent
In this video, we will walk through how to easily build a Typescript deep research agent This builds upon our new DeepAgents JS library. All it involves is some detailed prompting, some search too...
Build a Streaming LangChain Agent in Next.js with useStream
Learn how to build a minimal LangChain agent inside a Next.js app using the useStream hook. We’ll stream responses, render human/AI messages, and add conversation memory via a LangGraph checkpointe...
Human in the Loop Middleware (Python)
Learn about how to use LangChain's human in the loop middleware to approve, edit, and reject tool calls before they're executed. Our example uses an email assistant agent that requires human feed...
Why We Built LangSmith for Improving Agent Quality
Harrison Chase (CEO of LangChain) sits down with Bagatur (LangSmith Engineer) and Tanushree (Product Manager) for a technical roundtable on bringing production agents from prototype to rigor. They ...
Deep Agent CLI: Coding Assistant with Memory
Using the deepagents package, we built a simple coding CLI as an example of a coding application you could build on top of deepagents. We added in a concept of memory so that it would remember ins...
Inside LangSmith's No Code Agent Builder
Harrison Chase (CEO of LangChain) sits down with Brace (Applied AI) and Sam (PM) for a technical roundtable on LangChain's first no code agent builder. They share how business users and engineers a...
Get Started with LangSmith Agent Builder
LangSmith Agent Builder (now in private preview) lets anyone create agents through a text-to-agent experience — no coding or prompt engineering required. Just describe what you want in plain langua...
LangChain Academy New Course: LangGraph Essentials
In our newest LangChain Academy course, LangGraph Essentials, you can learn the basics of LangGraph in less than an hour in either Python or TypeScript. LangGraph is a low-level orchestration fram...
LangChain Academy New Course: LangChain Essentials
In our newest LangChain Academy course, LangChain Essentials, you can learn the basics of LangChain in less than an hour in either Python or TypeScript. LangChain is the best way to get started wi...
Get Started with LangSmith Multi-turn Evaluations
Once you have a good sense of the top usage patterns your agent is handling, you can start to drill into how each complete conversation is performing. Multi-turn Evals help you measure whether your...
Building LangChain and LangGraph 1.0
LangChain CEO Harrison Chase sits down with open source engineers Sydney, Hunter, and Will for an in-depth technical discussion on the major 1.0 releases of LangChain and LangGraph. The team explor...
LangChain: Engineer reliable agents
We're launching LangChain 1.0 and LangGraph 1.0 — and announcing our $125M Series B. Building reliable agents has traditionally been hard. What started as frameworks for LLM applications has evolv...