--- title: "Getting Started" id: "1828" type: "page" slug: "getting-started" published_at: "2026-09-20T23:47:09+00:00" modified_at: "2026-10-22T16:38:23+00:00" url: "https://xedant.com/agents/chatbot/docs/getting-started/" markdown_url: "https://xedant.com/agents/chatbot/docs/getting-started.md" excerpt: "Chatbot Agent is a program for building chatbots with artificial intelligence that runs on your…" --- # Getting Started [https://xedant.com/agents/chatbot/docs/getting-started/.md](https://xedant.com/agents/chatbot/docs/getting-started/.md) Chatbot Agent is a program for building chatbots with artificial intelligence that runs on your server. The dialog is assembled with the mouse in a visual studio, checked on the built-in emulator, and then the same bot answers in Telegram, on a public chat page and as a widget on your website. AI answers from your documents and honestly hands the conversation to a human when it cannot cope. ## What Chatbot Agent is It is not a cloud builder with a monthly subscription, but an application you run on your own server and fully control. A bot is a set of plain files: settings, the dialog scenario and scripts. You can see them, copy them to another server and move them without export or import. There is no lock-in to someone else’s service. From one scenario the bot can work in three channels at once: in Telegram, on a public chat page on the platform’s domain and as a widget on your website. The scenario is written once, and conversations from all channels meet in one place — the operator sees Telegram, the website and the widget together. ## Who it is for Small business owners, shops and support teams, professionals — everyone who needs a bot but not a programmer, and who does not want to pay for a cloud subscription. The scenario is assembled from ready-made block steps, and AI helps write the first draft from a plain description in words. When the blocks are not enough, you can connect your own Python script — for example, to look into your CRM or calculate an order. ## System requirements The product is deployed on a server through Xedant MultiAgent, as part of the Agent Factory subscription: Docker and everything else is installed by MultiAgent itself, and an inexpensive VPS will do. The application web server, the database and the Telegram emulator are already part of the product. ## First sign-in Open the application address in a browser. While the database has no users, the registration form shows by itself — the first person to register becomes the administrator. The second way is to set the login and password in advance with environment variables at startup (the variable stores not the password itself but its “fingerprint” — a lowercase SHA-256 hash). A sign-in session lasts 90 days, so you will not have to sign in every week. If you have many applications and users, sign-in can be set up through an external broker — then users enter Chatbot Agent with their credentials from the shared system. That is optional: without it, sign-in is ordinary. ## The interface All sections are gathered in the side menu: **Home**, **Bots**, **Emulator**, **Inbox**, **Contacts**, **Analytics**, **Knowledge**, **Models**, **Scripts**, **Health** and **Live** (live conversations right now). Below are the project files and data files. At the very bottom, when Xedant Agent is connected, active chats with the AI agent and the task queue appear. The flow studio, like an individual bot’s settings, opens from the bot list — it is not a separate menu item but the bot’s “inner” screen. What lives where is described on the pages below. ## First steps 1. **Create a bot** — enter a name and a token from @BotFather. The token is the bot’s secret access key; Telegram issues it. No real bot yet? No problem: create a test bot on the emulator, it works without Telegram. 2. **Assemble the first dialog** in the studio — drag a couple of block steps onto the canvas and connect them with an arrow. 3. **Check it on the emulator** — write to the bot as a fake person and see what it answers. 4. **Press “Start”** — the bot starts receiving messages. 5. **Connect the rest as you like** — the knowledge base, an operator, the public chat page and the widget on your website. ## Language, theme and font size The interface language is set by the brand at installation: under the Xedant brand the interface is English. The theme (dark or light), the font size and the notification sound volume are configured in the appearance section and saved for the specific user — everyone signs in with their own settings. ## A desktop shortcut The application can be installed like a regular program: the browser offers to add a shortcut to the desktop or the application menu, and Chatbot Agent will open in its own window, without an address bar. No app stores or installers needed. An internet connection is required. ## All documentation sections [Bots](/agents/chatbot/docs/bots) — creating, starting and stopping, versions and the publish checklist. [Flow Studio](/agents/chatbot/docs/studio) — building the dialog with the mouse. [Flow Nodes](/agents/chatbot/docs/nodes) — all 18 step types. [Emulator](/agents/chatbot/docs/emulator) — checking without Telegram. [Testing & Debugging](/agents/chatbot/docs/testing) — dry runs and tracing. [Inbox](/agents/chatbot/docs/inbox) — the operator’s work. [Knowledge Base](/agents/chatbot/docs/knowledge) — answers from your documents. [Models & Prompts](/agents/chatbot/docs/models) — configuring AI. [Scripts](/agents/chatbot/docs/scripts) — your own Python programs and new communication channels. [Contacts, Segments & Broadcasts](/agents/chatbot/docs/contacts) — working with the client base. [Analytics & Logs](/agents/chatbot/docs/analytics) — numbers and records. [Website Chat Room](/agents/chatbot/docs/chatroom) — the public chat page. [Website Widget](/agents/chatbot/docs/widget) — a chat panel on your site. [Chat with the AI Agent](/agents/chatbot/docs/agent) — edits in plain words. [← Back to the documentation index](/agents/chatbot/docs)