# Introduction
URL: /docs/runtimes/langgraph/tutorial/introduction

Build a stockbroker assistant with LangGraph and assistant-ui.

> For AI agents: a documentation index is available at [llms.txt](/llms.txt). Use `.md` for canonical markdown pages; `.mdx` is kept as a backwards-compatible alias on supported URL paths.

In this tutorial, we will build a stockbroker assistant using LangChain.js, LangGraph.js and assistant-ui.

We will go through the necessary steps to integrate assistant-ui with a LangGraph Cloud endpoint. Code snippets focus on the setup of the frontend, but we will highlight relevant sections of the backend code as well.

This agent leverages the following features:

- 🚄 Streaming of messages from LangGraph state to assistant-ui
- 💅 Rich text rendering using Markdown
- 🛠️ Generative UI: Mapping tool calls to tool UIs
- 🔁 Approval UI: Confirming tool calls before execution (human-in-the-loop)

## Prerequisites

- Node.js 24 or higher

## Final result

- Demo: <https://assistant-ui-stockbroker.vercel.app/>
- Source Code: <https://github.com/assistant-ui/assistant-ui-stockbroker>

## Get started

Begin Part 1 of the tutorial by [setting up the frontend](/docs/runtimes/langgraph/tutorial/part-1).