HybridAI Logo HybridAI
Log in Sign Up

 

HybridAI User Manual

Table of Contents

1. Introduction

HybridAI is an advanced AI chatbot system that integrates multiple AI models, emotional detection, and function calls to create a seamless user experience. It supports direct chatbot interactions, website automation, and integrations with various platforms such as WhatsApp, Slack, and Telegram.

2. System Overview

3. Setting Up HybridAI

3.1 Embedding the Chatbot

To integrate HybridAI into your website, use the following embedding script:

<script>
    window.chatbotConfig = {
        chatbotId: "YOUR_CHATBOT_ID", // Your unique chatbot ID
        chatbotServer: "https://hybridai.one" // Backend server domain
    };
</script>
<script src="https://hybridai.one/hai_embed.js?chatbotId=YOUR_CHATBOT_ID"></script>

This script loads the chat widget and connects it to the HybridAI server. You can find the exact script parameters on the integration page in your admin panel.

HybridAI Chatbot Widget Integration

Example of a HybridAI chatbot widget integrated into a website.

3.2 Configuring the Chatbot

Go to Admin Panel > General Settings to:

General Settings Configuration Interface

The General Settings interface showing bot configuration options including name, color, messages, and pricing settings.

Upon request, we also support self-hosted and specialized models such as Teuken, Occiglot, Dolphin, and more.

3.3 Widget Integration Modes

HybridAI offers four different modes for integrating your chatbot into your website. Each mode is optimized for specific use cases and can be selected based on your requirements.

1. Content-Embedded ChatBot

Integrate the chatbot directly into your website content. Perfect for documentation sites, knowledge bases, and product pages where context is important.

The bot is embedded within existing content and can interact with specific page elements. This integration enables a seamless user experience where the chatbot appears as a natural part of the page.

2. Chat Icon with Floating Widget

Display a chat icon that opens a floating chat widget when clicked. Ideal for customer support and general user engagement without taking up permanent space.

The widget appears as a floating window over the page content and can be opened or closed as needed. This solution is particularly popular for e-commerce websites and support portals.

3. Default Open Floating Widget

Show a floating chat widget that is open by default. Great for high-engagement scenarios where you want the chat to be immediately visible and accessible.

This option is particularly suitable for landing pages, onboarding processes, or situations where you want to encourage maximum user interaction. The chat is present from the start and invites interaction.

4. Headless Mode via API

Use your chatbot programmatically through our API. Perfect for custom integrations, mobile apps, or when you need full control over the user interface.

In headless mode, you have complete control over design and functionality. You can embed the chatbot in your own applications, develop mobile apps, or create complex workflows.

3.4 Widget Configuration Options

You can customize the appearance and behavior of your chat widget through various configuration options:

Basic Configuration

<script>
    window.chatbotConfig = {
        chatbotId: "YOUR_CHATBOT_ID",
        chatbotServer: "https://hybridai.one",
        initialGreeting: "Welcome! How can I help you today?",
        widgetColor: "#3498db",
        position: "right",  // Options: "right", "left", "center"
        widgetSize: "standard"  // Options: "standard", "compact"
    };
</script>
<script src="https://hybridai.one/hai_embed.js?chatbotId=YOUR_CHATBOT_ID"></script>

Advanced Configuration Options

Example of Advanced Configuration

<script>
    window.chatbotConfig = {
        chatbotId: "YOUR_CHATBOT_ID",
        chatbotServer: "https://hybridai.one",
        initialGreeting: "Hello! I'm your personal assistant. How can I help?",
        widgetColor: "#2ecc71",
        position: "left",
        widgetSize: "compact",
        defaultOpen: false,
        showAvatar: true,
        enableNotifications: true,
        customCSS: {
            borderRadius: "12px",
            boxShadow: "0 4px 20px rgba(0,0,0,0.15)"
        }
    };
</script>
<script src="https://hybridai.one/hai_embed.js?chatbotId=YOUR_CHATBOT_ID"></script>

Note: Visit the Admin Panel > Integration page to get your specific embedding code and test different demo variants.

3.5 Platform Integrations

HybridAI can be integrated into various platforms and systems, either through universal JavaScript embedding or through special integrations we provide for popular platforms.

Universal JavaScript Integration

The simplest method to integrate HybridAI into any website or platform is using our JavaScript embedding script:

<script>
    window.chatbotConfig = {
        chatbotId: "YOUR_CHATBOT_ID",
        chatbotServer: "https://hybridai.one"
    };
</script>
<script src="https://hybridai.one/hai_embed.js?chatbotId=YOUR_CHATBOT_ID"></script>

This method works on any platform that supports JavaScript, including:

WordPress Plugin Integration

For WordPress websites, we offer a comprehensive plugin that enables seamless integration:

The HybridAI Chatbot WordPress Plugin offers additional features such as:

Other Platform Integrations

Besides WordPress, we also support integrations with other popular platforms:

Messaging Platform Integrations

HybridAI seamlessly integrates with popular messaging and social media platforms:

You can test our WhatsApp integration on the homepage. Contact us if you want to use these integrations for your bot.

Automation Platforms

HybridAI integrates with popular automation and workflow platforms:

These integrations allow you to trigger actions in other systems based on chatbot interactions or feed external data into your chatbot.

API Integration for Developers

For developers, we offer various API endpoints for custom integrations:

Note: For special platform integrations or custom solutions, contact us at service@hybridai.one. We're happy to help with implementation.

3.6 Advanced Widget Customization (CSS Customization)

The HybridAI Widget v2 now supports advanced CSS customization, allowing you to fully match your brand identity without compromising core functionality.

Customization Methods

Method 1: External CSS File (Recommended for Advanced Customization)

Load a completely custom CSS file that overrides the default widget styles:

<script>
    window.chatbotConfig = {
        chatbotId: "YOUR_CHATBOT_ID",
        customCssUrl: "https://yourdomain.com/path/to/custom-widget-styles.css",
        useDefaultCss: false  // Optional: disable default CSS entirely
    };
</script>
<script src="https://hybridai.one/hai_embed.js?chatbotId=YOUR_CHATBOT_ID"></script>

Benefits:

Method 2: CSS Variables (Quick Theming)

Override specific CSS variables for quick customization:

<script>
    window.chatbotConfig = {
        chatbotId: "YOUR_CHATBOT_ID",
        // Color customization
        color_scheme: "#FF6B6B",
        
        // Size customization
        width: "450px",
        height: "650px",
        
        // Position customization
        buttonPosition: {
            bottom: "30px",
            right: "30px"
        },
        
        // Typography
        fontSize: "16px",
        
        // Theme
        theme: "dark"  // or "light"
    };
</script>
Method 3: Hybrid Approach

Use the default CSS with variable overrides plus a custom CSS file for specific modifications:

<script>
    window.chatbotConfig = {
        chatbotId: "YOUR_CHATBOT_ID",
        customCssUrl: "https://yourdomain.com/widget-overrides.css",
        color_scheme: "#6C63FF",
        width: "500px"
    };
</script>

Available CSS Variables

The widget provides 30+ CSS variables that you can customize. Here's the complete list:

:root {
    /* Primary colors */
    --hai-primary-color: #f5f6f8;
    --hai-primary-hover: #e8e9eb;
    --hai-text-color: #333333;
    --hai-text-color-inverse: #ffffff;
    
    /* Widget dimensions */
    --hai-button-size: 60px;
    --hai-window-width: 400px;
    --hai-window-height: 600px;
    --hai-window-min-width: 350px;
    --hai-window-min-height: 400px;
    --hai-window-max-width: 90vw;
    --hai-window-max-height: calc(100vh - 40px);
    
    /* Positioning */
    --hai-button-position-bottom: 20px;
    --hai-button-position-right: 20px;
    
    /* Border and shadows */
    --hai-border-radius: 12px;
    --hai-button-border-radius: 50%;
    --hai-box-shadow: 0 10px 30px rgba(0, 0, 0, 0.15);
    --hai-button-shadow: 0 4px 8px rgba(0, 0, 0, 0.2);
    
    /* Z-index layering */
    --hai-z-index: 9999;
    
    /* Animation timing */
    --hai-transition-duration: 0.2s;
    --hai-animation-duration: 1.5s;
    
    /* Title bar */
    --hai-titlebar-height: 40px;
    --hai-titlebar-bg: #2196F3;
    --hai-titlebar-text: #ffffff;
    
    /* Font sizes */
    --hai-font-size-base: 14px;
    --hai-font-size-title: 14px;
    --hai-font-size-controls: 16px;
    --hai-font-size-tooltip: 12px;
    
    /* Spacing */
    --hai-spacing-xs: 4px;
    --hai-spacing-sm: 8px;
    --hai-spacing-md: 10px;
    --hai-spacing-lg: 15px;
    --hai-spacing-xl: 20px;
    
    /* Dark theme overrides */
    --hai-dark-bg: #1e1e1e;
    --hai-dark-text: #e0e0e0;
    --hai-dark-hover: #2a2a2a;
}

Example: Corporate Branding

/* corporate-theme.css */
:root {
    --hai-primary-color: #003366;
    --hai-titlebar-bg: #003366;
    --hai-button-size: 70px;
    --hai-border-radius: 8px;
    --hai-font-size-base: 16px;
}

.hybridai-chat-button {
    background: var(--hai-primary-color) !important;
    box-shadow: 0 4px 12px rgba(0, 51, 102, 0.3) !important;
}

Best Practices

Important: CSS customization is completely optional. If you don't provide any customization options, the widget will work exactly as before with the default inline styles.

Basic Widget Configuration

You can customize various aspects of the widget by modifying the chatbotConfig object:

<script>
    window.chatbotConfig = {
        chatbotId: "YOUR_CHATBOT_ID",
        chatbotServer: "https://hybridai.one",
        initialGreeting: "Welcome! How can I help you today?",
        widgetColor: "#3498db",
        position: "right",  // Options: "right", "left", "center"
        widgetSize: "standard",  // Options: "standard", "compact"
        context: "User is on the product page for Premium Widget Pro"  // Optional context information
    };
</script>
<script src="https://hybridai.one/hai_embed.js?chatbotId=YOUR_CHATBOT_ID"></script>

Context Feature

The context parameter allows you to provide information about where the widget is running. This context information gets included in the welcome message sent to the LLM, helping the AI understand the user's current situation and provide more relevant responses.

Use Cases:
Examples:
// Product page context
window.chatbotConfig = {
    chatbotId: "your-chatbot-id",
    context: "User is viewing the Premium Widget Pro product page. This is our flagship product that costs $299 and includes lifetime support."
};

// Support page context
window.chatbotConfig = {
    chatbotId: "your-chatbot-id",
    context: "User is on the troubleshooting page for connection issues. They may need help with network configuration."
};

// Shopping cart context
window.chatbotConfig = {
    chatbotId: "your-chatbot-id",
    context: "User is in the shopping cart with 3 items: Premium Widget Pro ($299), Extended Warranty ($49), and Free Shipping. Total cart value: $348."
};

Note: The context is optional. If not provided, the welcome message works as before without any context information.

Demo Page

To explore different widget configurations and test appearance options, visit our Widget Demo Page. This interactive demo allows you to:

3.7 Fullscreen Mode Customization

The HybridAI widget includes a fullscreen mode that provides a ChatGPT-like interface for users. This mode can be fully customized to match your brand identity and provide a seamless user experience.

Accessing Fullscreen Mode

Users can enter fullscreen mode by clicking the fullscreen button (⛶) in the widget's header. This opens a immersive chat interface that covers the entire viewport.

Customization Options

You can customize every aspect of the fullscreen mode through the fullscreen configuration object:

<script>
    window.chatbotConfig = {
        chatbotId: "YOUR_CHATBOT_ID",
        fullscreen: {
            // Background and Layout
            overlayBg: '#1a1a2e',           // Fullscreen background color
            headerBg: '#16213e',            // Header background
            headerBorderColor: '#0f3460',   // Header border color
            headerPadding: '20px 30px',     // Header padding
            contentMaxWidth: '900px',       // Maximum content width
            contentPadding: '0 30px',       // Content side padding
            
            // Message Styling
            messageMaxWidth: '85%',         // Maximum message width
            messagePadding: '15px 20px',    // Message padding
            messageBorderRadius: '12px',    // Message border radius
            messageFontSize: '17px',        // Message font size
            userMessageBg: '#4a90e2',       // User message background
            userMessageColor: '#ffffff',    // User message text color
            botMessageBg: '#2c3e50',        // Bot message background
            botMessageColor: '#ecf0f1',     // Bot message text color
            
            // Input Area
            inputPadding: '25px 0 50px 0',  // Input area padding
            inputBorderRadius: '15px',      // Input border radius
            inputFontSize: '17px',          // Input font size
            inputBorderColor: '#52687a',    // Input border color
            inputTextColor: '#ffffff',      // Input text color
            inputPlaceholderColor: '#95a5a6', // Placeholder text color
            
            // Welcome Text
            welcomeTitleSize: '52px',       // Welcome title font size
            welcomeSubtitleSize: '20px',    // Welcome subtitle font size
            welcomeTitleColor: '#ffffff',   // Welcome title color
            welcomeSubtitleColor: '#bdc3c7', // Welcome subtitle color
            
            // Buttons
            buttonBg: '#e74c3c',            // Button background
            buttonColor: '#ffffff',         // Button text color
            buttonHoverBg: '#c0392b'        // Button hover background
        }
    };
</script>

Alternative: External CSS Override

For advanced customization, you can use external CSS to override the fullscreen mode styling:

<style>
    :root {
        --fs-overlay-bg: #1a1a2e !important;
        --fs-header-bg: #16213e !important;
        --fs-user-message-bg: #4a90e2 !important;
        --fs-bot-message-bg: #2c3e50 !important;
        --fs-welcome-title-color: #ffffff !important;
    }
    
    .hybridai-fullscreen-overlay {
        font-family: 'Your Custom Font', sans-serif !important;
    }
</style>

Available CSS Variables

The fullscreen mode supports the following CSS variables that can be overridden:

Note: Fullscreen customization is completely optional. If you don't provide any fullscreen configuration, the widget will use default styling that adapts to your chosen theme (light/dark).

4. AI Settings

HybridAI offers a comprehensive selection of state-of-the-art AI models, each optimized for different use cases and requirements. Our platform supports leading providers including OpenAI, Google, Mistral, and more, with both cloud-based and self-hosted options available.

4.1 AI Model Selection

Choose from our carefully curated selection of AI models, each with distinct advantages:

AI Settings Configuration Interface

The AI Settings interface showing instruction editing, optimization tools, model selection, and temperature control.

OpenAI Models - Industry Leading Capabilities

GPT-4 and GPT-3.5 Turbo represent the current gold standard in conversational AI:

Google Gemini - Outstanding Performance

Gemini Pro and Gemini Pro Vision deliver exceptional performance with advanced capabilities:

Mistral Models - European Union Stack

Mistral Large and Mistral Medium provide EU-compliant AI solutions with excellent performance:

Specialized and Self-Hosted Models

For specific requirements, we also support:

Model Selection Guidelines

Choose your AI model based on your specific needs:

4.2 Instruction Optimization

The instruction system is the foundation of your chatbot's personality, knowledge, and behavior. Well-crafted instructions are crucial for creating effective, engaging, and helpful AI assistants.

Why Instructions Matter

Your chatbot's instructions serve as its "DNA" - they determine:

HybridAI's Instruction Optimization Tools

HybridAI provides powerful tools to help you create and optimize effective instructions:

1. AI-Powered Optimization

Use the "Optimize" button to enhance your instructions automatically:

2. Website-Based Instruction Generation

Use the "Use URL" function to automatically derive instructions from your website:

Best Practices for Effective Instructions

Follow these guidelines to create powerful instructions:

Structure and Organization
Content Elements

Instruction Optimization Workflow

Follow this process to create optimal instructions:

  1. Start with URL: Use your website URL to generate initial instructions
  2. Review and Customize: Modify the generated content to match your specific needs
  3. Add Specifics: Include unique business information, procedures, or policies
  4. Optimize: Use the "Optimize" button to refine and improve
  5. Test and Iterate: Test with real conversations and refine based on performance
  6. Regular Updates: Keep instructions current as your business evolves

Pro Tip: The most effective instructions are specific, clear, and comprehensive. Don't hesitate to include detailed information about your business, processes, and user expectations. The AI performs better with more context rather than less.

4.3 Additional AI Configuration

Beyond model selection and instructions, the AI Settings panel offers additional options to:

4.4 Prompt Plugins - Dynamic Personality Modules

Prompt Plugins allow your chatbot to automatically switch between different personalities or specialized instruction sets based on conversation context. This powerful feature enables your bot to adapt its expertise and communication style dynamically without user intervention.

Understanding Prompt Plugins

When activated, a plugin completely replaces your bot's base instructions with specialized instructions tailored for specific scenarios. For example:

Configuring Prompt Plugins

To set up prompt plugins for your chatbot:

  1. Navigate to Admin Panel > AI Settings > Prompt Plugins
  2. Click "Add New Plugin" to create a plugin
  3. Configure the following settings:
    • Plugin Name: A descriptive name for the plugin (e.g., "Conflict Resolver")
    • Plugin Prompt: The complete replacement instructions that define the bot's new personality and behavior
    • Activation Context: Description of when this plugin should activate (e.g., "When users discuss conflicts, disagreements, or interpersonal problems")
    • Activation Threshold: Confidence level (0.0-1.0) required for activation. Higher values require stronger context matching
  4. Click "Save Plugin" to add it to your bot's configuration

How Plugin Activation Works

The plugin system uses intelligent context analysis:

  1. Each incoming message is analyzed against all active plugins' activation contexts
  2. The system evaluates how well the message matches each plugin's context (0-1 confidence score)
  3. If a plugin's confidence exceeds its threshold, it activates
  4. When multiple plugins qualify, the one with highest confidence wins
  5. The bot's base instructions are completely replaced with the plugin's instructions
  6. The plugin remains active for the duration of that conversation session

Testing and Managing Plugins

The Prompt Plugins interface provides tools for testing and management:

Best Practices for Prompt Plugins

To get the most out of prompt plugins:

Note: Prompt plugins are evaluated using OpenAI's GPT-4 for consistent, high-quality context analysis, regardless of which model you use for chat responses. This ensures reliable plugin activation across all supported LLM providers (OpenAI, Google Gemini, Mistral, etc.).

5. Knowledge Base Extension

HybridAI allows you to extend your chatbot's knowledge base through document uploads and web crawling, so your bot is better informed about topics relevant to your business or use case.

5.1 Document Upload

In the Admin Panel > AI Settings > Chat & Knowledge section, you can upload various document types:

Uploaded documents are processed and converted into a format that allows the AI to reference and use this knowledge when answering user questions. The system automatically extracts content while preserving the structure and relationships between different parts of the document.

5.2 Web Crawling

HybridAI offers a powerful web crawling feature that allows your chatbot to learn from website content:

  1. Navigate to Admin Panel > AI Settings > Chat & Knowledge
  2. Find the Web Crawling section
  3. Enter the base URL from which the bot should learn (e.g., "https://yourdomain.com/knowledge/")
  4. Set the maximum number of pages to crawl (recommended: start with 10-20 pages)
  5. Click "Start Crawling" to begin the process

The system automatically explores the website, follows links from the specified URL, and extracts content that is incorporated into your chatbot's knowledge base. The crawling progress is displayed in real-time, showing which pages are currently being processed.

Note: For best results, direct the crawler to specific sections of your website that contain the most relevant information. Avoid crawling very large websites at once by starting with the most important sections.

5.3 Knowledge Management

After uploading documents or crawling websites, you can manage your knowledge base:

The AI will automatically incorporate this knowledge into conversations when relevant questions are asked.

6. Image Integration

HybridAI supports image integration, which allows your chatbot to display relevant images during conversations to improve user engagement and information delivery.

6.1 Uploading Images

To add images to your chatbot's library:

  1. Go to Admin Panel > AI Settings > Chat & Knowledge
  2. Scroll to the Upload Image section
  3. Select an image file (supported formats: PNG, JPG, JPEG, GIF, SVG, WEBP)
  4. Add a descriptive caption that explains what the image shows
  5. Click "Upload Image" to add it to your bot's library

6.2 How Image Integration Works

Images are automatically integrated into your chatbot's instruction system. The bot understands:

When a user asks a question or discusses a topic related to one of your uploaded images, the chatbot can choose to display the image alongside its text response to improve understanding.

6.3 Managing Images

You can manage your image library through the same interface:

7. Website Commands

Website commands allow the chatbot to dynamically interact with elements on the web page. These can highlight sections, fill forms, navigate, or change page content.

7.1 How Website Commands Work

When a chatbot recognizes a relevant action, it triggers a JavaScript event that modifies the web page.

To activate this, you need to add it to the bot's instructions, e.g., When asked about "Scope 1", include the command [CMD:highlight=Scope1] in your response.

Website commands are an integral part of the website bot (located at the bottom right of the website), you just need to activate them by making them part of the agent's instructions (see example below).

Website commands target HTML elements via their ID property. Please use properties in lowercase. Avoid spaces in the ID property and use underscores instead.
If a command cannot be executed, you can check the browser console to see which element the system is trying to address.

7.2 Available Commands

CommandDescription
highlightElement(selector)Highlights a web page element.
fillInput(selector, value)Fills an input field with a predefined value.
navigateTo(url)Redirects the user to another web page.

8. Human Takeover

HybridAI includes a human takeover feature that allows human agents to seamlessly intervene in a conversation when needed.

8.1 Triggering a Takeover

8.2 Admin Panel for Takeover

In the Admin Panel > Chat Console, administrators can:

8.3 Sensing & Trigger

In Sensing & Trigger, you can adjust the threshold at which a sensing dimension begins to activate the notification.

It's also possible to configure two custom dimensions that describe relevant events for your use case.

8.4 Alert Controls (Admin Chat Console)

In the Admin Chat Console, you can fine-tune how alerts surface to operators:

Thresholds are configured in Sensing & Trigger. Alerts are generated when any metric meets or exceeds its configured threshold.

9. AI Functions & API Integration

9.1 Function Calls

HybridAI supports function calls to extend chatbot capabilities. You can activate predefined functions such as:

You can try the currently integrated functions found on the function call page. Contact us if you want to implement your own functions. For security reasons, this will be available later in self-service mode.

9.2 Bot Memory API (IoT & External Data Integration)

HybridAI offers a Bot Memory API that allows you to transfer external data (such as IoT measurements, user preferences, or context information) directly into your bot's memory. This enables your bot to have persistent knowledge that improves its responses.

Key Features:

API Endpoints:

Store/update memory:

POST /v1/bot/memory
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

{
    "bot_id": "YOUR_BOT_ID",
    "memory_key": "temperature_sensor_1",
    "memory_content": "Current temperature: 22.5°C, humidity: 65%",
    "memory_type": "sensor_data"  // optional, default: "general"
}

Retrieve memory:

GET /v1/bot/memory?bot_id=YOUR_BOT_ID&memory_key=temperature_sensor_1
Authorization: Bearer YOUR_API_KEY

Delete memory:

DELETE /v1/bot/memory?bot_id=YOUR_BOT_ID&memory_key=temperature_sensor_1
Authorization: Bearer YOUR_API_KEY

Use Cases:

Example: IoT Temperature Monitoring

import requests

# Your IoT device sends temperature data
api_url = "https://hybridai.one/v1/bot/memory"
headers = {"Authorization": "Bearer YOUR_API_KEY"}

# Update temperature measurement
data = {
    "bot_id": "YOUR_BOT_ID",
    "memory_key": "office_temperature",
    "memory_content": "Office temperature at 2:30 PM: 23.5°C (comfortable range)",
    "memory_type": "iot_sensor"
}

response = requests.post(api_url, json=data, headers=headers)

# When users now ask "What's the office temperature?", 
# the bot has this information in its context
Note: The Bot Memory API uses the same API key as the OpenAI-compatible API. Generate your API key in the Admin Panel > API Keys section.

9.3 OpenAI-Compatible API

HybridAI now offers an OpenAI-compatible API that allows you to integrate your custom chatbots with your applications. Key features include:

Using the API

To use the API, first generate an API key in the Admin Panel > API Keys section. Then use the API with standard libraries:

import openai

# Configure with your API key and HybridAI base URL
openai.api_key = "YOUR_API_KEY"
openai.base_url = "https://hybridai.one/"

# Make a chat completion request
response = openai.chat.completions.create(
    model="gpt-3.5-turbo",  # You can use available models
    messages=[
        {"role": "user", "content": "Hello, world!"}
    ],
    chatbot_id="YOUR_BOT_ID"  # Specify which of your bots you want to use
)

print(response.choices[0].message.content)

9.4 N8N Integration

HybridAI offers native integration with N8N, a powerful workflow automation platform. This integration allows you to incorporate complex automations and workflows directly into your chatbots through natural conversation.

What is N8N?

N8N is an open-source workflow automation platform that enables you to connect various services and applications. With N8N, you can:

How N8N Integration Works

The HybridAI N8N integration transforms your N8N webhooks into AI function calls that can be invoked naturally through conversation. Here's how it works:

  1. Webhook Discovery: The system automatically scans your N8N workflows for webhook triggers
  2. Function Creation: Convert discovered webhooks into AI function calls with custom parameters
  3. Natural Invocation: Users simply ask the chatbot to perform actions using natural language
  4. Automatic Execution: The AI identifies the intent and calls the appropriate N8N workflow with the correct parameters

Setting Up N8N Integration

To set up N8N integration for your chatbot:

  1. Go to Admin Panel → AI Settings → N8N Integration
  2. Enter your N8N API key (get it from n8n.hybridai.one)
  3. Set the N8N base URL (default: https://n8n.hybridai.one)
  4. Click "Discover Webhooks" to scan your active workflows
  5. For each webhook, click "Create Function Call" to convert it into an AI function
  6. Define function parameters using JSON schema format
  7. Test the integration through natural conversation
N8N Integration Configuration Interface

The N8N Integration interface showing API configuration, webhook discovery, and function call management.

Creating Function Calls from Webhooks

When creating function calls from N8N webhooks, you need to define:

Example Parameter Schema

{
  "type": "object",
  "properties": {
    "message": {
      "type": "string",
      "description": "The notification message to send"
    },
    "recipient": {
      "type": "string", 
      "description": "Email address or username to notify"
    },
    "priority": {
      "type": "string",
      "enum": ["low", "normal", "high"],
      "description": "Priority level of the notification"
    }
  },
  "required": ["message", "recipient"]
}

Natural Language Usage

Once N8N functions are set up, users can invoke them naturally:

Use Cases

Popular use cases for N8N integration include:

Example: Customer Support Automation

A practical example of N8N integration for customer support:

  1. User: "I'm having trouble with my account login"
  2. AI: Automatically calls `create_support_ticket` function
  3. N8N Workflow: Creates ticket in support system, assigns to appropriate team
  4. AI Response: "I've created support ticket #12345 for your login issue. Our team will contact you within 24 hours."
  5. Additional Actions: N8N can send confirmation email, notify support team, etc.

Native N8N Instance

We provide a native N8N instance at n8n.hybridai.one, specifically configured for integration with HybridAI chatbots. This instance offers:

Getting Started: Access your N8N instance at n8n.hybridai.one using your HybridAI credentials. The N8N integration feature is available in the Admin Panel under AI Settings → N8N Integration.

10. Analytics & Logs

HybridAI provides real-time analytics and logs for chatbot interactions. Simply visit the "Analytics" page to access:

13. Multi-Bot Management

HybridAI allows you to create and manage multiple chatbots under one account, each with its own configuration:

14. Business Intent Extraction

HybridAI includes an advanced Business Intent Extraction feature that automatically identifies and extracts business-relevant information from customer conversations using AI analysis.

14.1 Overview

The Business Intent Extraction system analyzes user messages in real-time to identify important business information such as:

14.2 Configuration

To configure business intent extraction:

  1. Navigate to Admin Panel → AI Sensing & Triggers
  2. Scroll down to the Business Intent Extraction section
  3. Click Add Intent Type to create new extraction rules
  4. Define your custom entity types with descriptive names
  5. Provide clear descriptions to guide the AI extraction
  6. Enable/disable specific entity types as needed
  7. Click Save Intent Configuration

Entity Configuration Examples

14.3 Viewing Extracted Data

To view and analyze extracted business intents:

  1. Navigate to Admin Panel → Analytics
  2. Access the Business Intents Analytics page at /admin_analytics_intents
  3. Use the filtering options to refine your view:
    • Date Range: Filter by specific time periods
    • Entity Type: Focus on specific types of extracted data
  4. Export data to CSV for further analysis

14.4 Data Analysis & Export

The Business Intent Analytics dashboard provides:

14.5 Integration with Emotional KPIs

Business intent extraction works seamlessly with existing emotional KPI analysis:

14.6 Use Cases

Popular applications for business intent extraction include:

Note: Business intent extraction requires active configuration to function. By default, no entities are extracted until you define custom entity types in the AI Sensing settings.

15. Contact & Support

For support, contact service@hybridai.one or visit the HybridAI Website.

Last updated: 19.05.2025