> ## Documentation Index
> Fetch the complete documentation index at: https://docs.junis.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction to Junis

> Build powerful multi-agent AI systems without writing code

## Welcome to Junis

Junis is an AI workforce platform where organizations build, orchestrate, and deploy teams of AI agents without writing code. Each team combines 4 agent types (LLM, Sequential, Parallel, Loop) with MCP integrations, 200+ Composio OAuth apps, a shared organization wiki, persistent memory, an embeddable chat widget, and an OpenAI-compatible API.

***

## What is Junis?

Junis transforms how you build with AI by providing:

<CardGroup cols={2}>
  <Card title="Multi-Agent System" icon="network-wired">
    Orchestrate multiple AI agents that collaborate to solve complex tasks
  </Card>

  <Card title="No-Code Interface" icon="wand-magic-sparkles">
    Build sophisticated AI workflows through an intuitive visual interface
  </Card>

  <Card title="Enterprise-Ready" icon="building">
    Multi-tenant architecture with organization-level isolation and security
  </Card>

  <Card title="Flexible Integration" icon="plug">
    Connect to external tools, databases, and APIs seamlessly
  </Card>
</CardGroup>

***

## What are Junis's key features?

### 🤖 Dynamic Agent System

Create four types of agents to handle any workflow:

| Agent Type           | Purpose                                     | Use Case                                            |
| -------------------- | ------------------------------------------- | --------------------------------------------------- |
| **LLM Agent**        | Executes AI tasks using language models     | Chat, analysis, content generation                  |
| **Sequential Agent** | Runs agents in order, step by step          | Multi-stage workflows (research → analyze → report) |
| **Parallel Agent**   | Runs multiple agents simultaneously         | Data collection from multiple sources               |
| **Loop Agent**       | Repeats agent execution until condition met | Iterative refinement, quality checks                |

### 🔧 Powerful Integrations

<AccordionGroup>
  <Accordion title="MCP Integration" icon="link">
    Connect to external services including GitHub and Firecrawl via Model Context Protocol. Add custom MCP servers for your internal tools.
  </Accordion>

  <Accordion title="RAG Knowledge Base" icon="book">
    Upload documents and create searchable knowledge bases. Your agents can query PDFs, Word docs, spreadsheets, and more.
  </Accordion>

  <Accordion title="Custom Tools" icon="hammer">
    Register Python functions as agent tools. Call APIs, process data, or integrate with any service.
  </Accordion>

  <Accordion title="Web Search" icon="magnifying-glass">
    Enable real-time web search for agents to access the latest information beyond their training data.
  </Accordion>
</AccordionGroup>

### 🧠 Second Brain & Collaboration

<CardGroup cols={2}>
  <Card title="Organization Wiki" icon="book-open" href="/guides/wiki-second-brain">
    A shared second brain every agent reads and writes — say "save this to the wiki" once and your whole AI team knows it
  </Card>

  <Card title="Squad Collaboration" icon="people-arrows" href="/guides/orchestration/squad-collaboration">
    One squad's orchestrator asks another squad a question and relays the answer — chain depth 3, loops blocked
  </Card>

  <Card title="Widget Embed" icon="window-maximize" href="/guides/widget-embed">
    Put your AI team on any website with a single iframe — visitors chat without logging in
  </Card>

  <Card title="B2B2C API" icon="code" href="/api-reference/user-provisioning">
    OpenAI-compatible API plus user provisioning to build your own product on top of Junis
  </Card>
</CardGroup>

### 💬 Seamless Chat Experience

* **Real-time Streaming**: See agent responses as they're generated
* **Multi-turn Conversations**: Maintain context across multiple messages
* **File Attachments**: Upload images, documents, and files for analysis
* **Voice Input**: Record and transcribe voice messages

### 💳 Flexible Billing

<CardGroup cols={2}>
  <Card title="Subscription Plans" icon="credit-card">
    Monthly plans with included credits for AI agent usage
  </Card>

  <Card title="Credit System" icon="coins">
    Pay-per-use credits for flexible consumption tracking
  </Card>
</CardGroup>

Manage your subscription and monitor usage from the **Subscription** page. Credits are automatically deducted based on agent activity.

### 🔐 Enterprise Security

<CardGroup cols={2}>
  <Card title="Organization Isolation" icon="building-shield">
    Each organization has its own agents, data, and configurations
  </Card>

  <Card title="Role-Based Access" icon="user-shield">
    Fine-grained permissions: Super Admin, Admin, Support, Member
  </Card>

  <Card title="API Key Management" icon="key">
    Generate and manage API keys for external integrations
  </Card>

  <Card title="Audit Logging" icon="clipboard-list">
    Track all actions and changes for compliance
  </Card>
</CardGroup>

***

## How does Junis work?

```mermaid theme={null}
graph LR
    A[User Message] --> B[Orchestrator]
    B --> C{Agent Selection}
    C --> D[LLM Agent]
    C --> E[Sequential Agent]
    C --> F[Parallel Agent]
    D --> G[Tools & MCP]
    E --> G
    F --> G
    G --> H[Response]
    H --> A
```

1. **User sends a message** through the chat interface or API
2. **Orchestrator analyzes** the request and selects the appropriate agent
3. **Agent executes** using LLMs, tools, RAG, or MCP integrations
4. **Response streams** back to the user in real-time
5. **Context persists** across the conversation session

***

## Who Uses Junis?

<Tabs>
  <Tab title="Product Teams">
    Build customer support chatbots, onboarding assistants, and internal tools without engineering resources.
  </Tab>

  <Tab title="Data Teams">
    Automate data analysis, generate reports, and query multiple databases simultaneously.
  </Tab>

  <Tab title="Content Creators">
    Generate blog posts, social media content, and marketing copy with multi-agent orchestrations.
  </Tab>

  <Tab title="Developers">
    Integrate AI capabilities into your apps using OpenAI-compatible APIs and webhooks.
  </Tab>
</Tabs>

***

## How does Junis compare to other platforms?

| Feature                 | Junis                        | LangChain          | AutoGPT         |
| ----------------------- | ---------------------------- | ------------------ | --------------- |
| **No-Code Interface**   | ✅ Visual workflow builder    | ❌ Code-based       | ❌ Code-based    |
| **Multi-Agent System**  | ✅ 4 agent types              | ✅ Limited          | ✅ Limited       |
| **Real-time Streaming** | ✅ WebSocket + SSE            | ⚠️ Basic           | ❌ Polling       |
| **Multi-Tenant**        | ✅ Organization isolation     | ❌ Self-hosted only | ❌ Single user   |
| **MCP Integration**     | ✅ GitHub, Firecrawl + Custom | ❌ Manual setup     | ❌ Not supported |
| **API Compatibility**   | ✅ OpenAI-compatible          | ❌ Custom           | ❌ Custom        |
| **Deployment**          | ✅ Cloud + On-premise         | ⚠️ Self-hosted     | ⚠️ Self-hosted  |

***

## Quick Start

Ready to build your first AI agent?

<CardGroup cols={2}>
  <Card title="5-Minute Tutorial" icon="rocket" href="/getting-started/quickstart">
    Create your first chatbot in 5 minutes
  </Card>

  <Card title="Agent System Guide" icon="brain" href="/getting-started/agent-system">
    Learn about the four agent types
  </Card>

  <Card title="API Documentation" icon="code" href="/api-reference/authentication">
    Integrate Junis into your applications
  </Card>
</CardGroup>

***

## Support & Community

<CardGroup cols={2}>
  <Card title="Documentation" icon="book" href="/getting-started/quickstart">
    Comprehensive guides and tutorials
  </Card>

  <Card title="Email Support" icon="envelope" href="mailto:contact@junis.ai">
    Contact our team for assistance
  </Card>
</CardGroup>

***

<Note>
  **New to AI agents?** Don't worry! Our [Quick Start Guide](/getting-started/quickstart) walks you through creating your first agent step-by-step, with no prior knowledge required.
</Note>