Chatbot Submenu
The Chatbot submenu contains all configuration settings for the AI-powered bot. Here you define how the bot behaves, which documents it uses as a knowledge source, and how it responds to users. After completing all configurations, you can test the bot in the Playground environment.
The Chatbot submenu consists of three main sections:
Documents
Agents
AI Testing
1. Documents
Overview
The Documents section allows users to build a knowledge base. Users can upload .docx and .md files or add HTML pages via URL. All uploaded content appears in the right panel.
1.1 File Upload
Supported formats:
.docx
.md
Limitations:
Maximum size: 10MB per file
Drag & drop supported
Multiple files can be uploaded at once
Process:
User selects or drags a file
Backend validates format and size
File is stored
Metadata (name, type, date) is saved
Document appears in the right panel
1.2 URL Upload
Users can enter a webpage URL and save it as a document.
Supported:
HTML pages
Blog posts
Online documentation
Limitations:
Maximum 5 URLs
Only valid URLs allowed
Duplicate URLs are not saved
Process:
URL is entered
Backend fetches the page
Script and style elements are removed
Clean content is stored
Appears as type HTML in the right panel
1.3 Uploaded Documents (Right Panel)
Each item displays:
Name or URL
Date and time
Type (DOCX / MD / HTML)
Delete option (update optional)
Purpose:
Centralized knowledge base management.
2. Agents
Overview
Agents are AI assistants designed to handle specific types of conversations. Each agent is configured with its own behavior and response categories.
2.1 Create Agent
A new agent is created via the Create Agent button.
Required fields:
Name – agent identifier
When to use – conditions or intent for selection
Behavior description – tone, priorities, rules
Response categories – assigned categories
Result:
The agent is added to the list.
2.2 System Instruction
After creation, a system instruction is assigned.
Defines:
Exact response rules
Behavior constraints and priorities
Once saved, the agent is fully configured.
2.3 Agent Management
Right-click options:
Edit
Activate / Deactivate
Delete
Statuses:
Active
Inactive
3. AI Testing
Overview
AI Testing is a real-time environment for testing agents before production.
Structure
Conversations Panel (Left)
Create new chat
View chat history
Date and time displayed
Chat Window (Center)
User messages on the right
AI responses on the left
Agent name and timestamps shown
Input Field (Bottom)
Enter message
Send message
Logic
Agent responds based on system instruction
Works within assigned categories
Can use the document knowledge base
Rejects out-of-scope requests
Purpose
Pre-production testing environment for validation and optimization of agents.