AI Chatbot Setup

Summary: The AI Chatbot within the Antimanual ecosystem provides sophisticated, real-time support by leveraging your indexed knowledge base. This guide covers everything from configuring the bot’s identity and conversational behavior to setting up lead capture, managing visual visibility, and implementing robust security measures to ensure a professional, secure user experience throughout.

The AI Chatbot within the Antimanual ecosystem serves as a sophisticated front-line engagement tool, capable of providing real-time support and gathering valuable user data. By leveraging your indexed and knowledge base, the chatbot delivers context-aware responses that reflect your brand’s unique identity and technical requirements.

Table of Contents

• Basic Configuration and Identity
• Conversational Behavior and Rules
• Lead Capture and Engagement
• Widget Display and Visibility
• Content Moderation and Security
• Frequently Asked Questions

Basic Configuration and Identity

The initial setup phase focuses on the external-facing persona of your AI assistant. This identity is crucial for establishing trust and setting expectations for the user interaction.

• Enable Chatbot: A global toggle to activate or deactivate the chat widget on your website’s frontend.
• Bot Name: The public name displayed in the header of the chat window (e.g., “Support Specialist” or “DocBot”).
• Bot Persona: A brief internal description that helps define the AI’s role, such as “Technical Documentation Assistant” or “Friendly Sales Representative.”
• Welcome Message: The automated greeting triggered when a visitor first opens the widget.
• Avatar: A custom image or brand logo that serves as the profile picture for the AI agent.

Conversational Behavior and Rules

To ensure the AI provides accurate and helpful information, you must define the logic and constraints of its communication style. This is primarily managed through the settings that feed the AI’s logic.

• AI Persona/Instruction: These are high-level system instructions. For example: “You are a senior developer support agent. Use technical terms where appropriate but remain polite. If you do not know the answer based on the knowledge base, direct the user to open a support ticket.”
• Pre-built Questions: Suggested prompts that appear as clickable buttons for users, allowing them to start common inquiries with a single click.
• Language Support: Selection of the primary language for communication to ensure the AI utilizes the correct linguistic nuances.
• Response Length: Settings to control verbosity, ranging from “Concise” (brief answers) to “Detailed” (comprehensive explanations).

Lead Capture and Engagement

For Pro users, the chatbot acts as a lead generation engine. By requiring contact information before or during a session, you can transform anonymous visitors into qualified leads.

• Lead Capture: An optional field to collect user emails. This is particularly effective when the AI cannot answer a question and offers to have a human follow up.
• Offline Message: The text shown to users when the AI services are unavailable or outside of specified operating hours.
• Success Message: A confirmation note displayed after a user successfully submits their lead information.

Widget Display and Visibility

Visual consistency and strategic placement are vital for a high-quality user experience. The display settings allow you to integrate the widget seamlessly into your site’s design.

• Widget Position: Options to dock the chat bubble on the bottom-left or bottom-right of the viewport.
• Theme Color: Customization of primary and secondary colors to match your corporate branding.
• Visibility Rules: Granular controls to show the bot only on specific pages (e.g., Documentation or Pricing) or to hide it entirely on mobile devices to save screen real estate.

Content Moderation and Security

Maintaining a professional environment requires strictly defined boundaries. The advanced security settings prevent the AI from engaging in off-topic or inappropriate conversations.

• Blocked Words/Phrases: A comma-separated list of prohibited terms. If a user inputs these, the AI will refuse to process the query.
• Block Response Message: A customized fallback message provided to the user when a moderation filter is triggered (e.g., “I am sorry, but I cannot discuss that topic.”).

Frequently Asked Questions

Can the chatbot pull data from my PDFs?
Yes, if you have uploaded and indexed your documents via the section, the chatbot will use that content to answer queries.

How do I change the chatbot’s icon?
Navigate to the “Display & Embedding” tab where you can choose from pre-set icons or upload a custom SVG/PNG file for your launcher button.

Does the chatbot work on mobile?
By default, yes. However, you can use the Visibility Rules to disable the widget for mobile users if you find it interferes with your site’s mobile responsiveness.