Chatbots

Chatbots are AI-powered conversational interfaces that can answer questions using your datasets.

Creating a Chatbot

1. Navigate to My Chatbots
2. Click Create Chatbot
3. Configure your chatbot:
   • Name: A descriptive name
   • System Prompt: Instructions for how the AI should behave
   • Datasets: Connect one or more datasets
   • Type: Private, Shared, or Public
4. Click Create

System Prompts

The system prompt defines your chatbot's personality and behavior. Good system prompts:

• Define the chatbot's role (e.g., "You are a helpful product expert")
• Set boundaries (e.g., "Only answer questions about our products")
• Specify tone (e.g., "Be friendly and concise")
• Provide context (e.g., "You have access to our product catalog")

Example

You are a knowledgeable assistant for Village Data.
Answer questions about datasets, chatbots, and AI features.
Be helpful and concise. If you don't know something, say so.

Connecting Datasets

Chatbots use RAG (Retrieval-Augmented Generation) to find relevant information:

1. User asks a question
2. System searches connected datasets for relevant content
3. Relevant content is provided to the AI as context
4. AI generates a response using that context

You can connect multiple datasets to a single chatbot.

Visibility Types

Type	Description
Private	Only you can access
Shared	Invite specific users
Public	Anyone can access (when published)

Publishing

To make a chatbot available to others:

1. Set type to Public (or Shared for specific users)
2. Change status from Draft to Published

See Content Visibility for detailed rules.

Usage Analytics

Published chatbots track usage metrics:

• Total conversations
• Message count
• User engagement

These help you understand how your chatbot is being used.