Chatbot Configuration Guide

Open Bot Identity

Pick your chatbot in the dashboard sidebar's chatbot menu, then open Configurations → Bot Identity. This page controls core personality and knowledge shared across every configured channel.

Set Bot Name and Model

Bot Name is required. Use a recognizable internal or customer-facing name such as Customer Support Bot.

Use Model Selection to choose the model that powers responses. Models can differ in price and capability. Select Rates beside the field to compare current pricing; do not rely on model names or prices copied from older documentation.

Optional: Select a fallback model, this model will take over when there is a disruption of service. Best practice is to select a fallback model that is of a different provider to the primary model.

Language

The bot can in the language you defined or response in the language a user used. Setting language to auto and the bit will adapt to the language used.

Write Bot Intro

Bot Intro is the greeting shown to a new visitor. Keep it short, explain what the bot can help with, and match your brand voice.

Example: Hi! I can help with product questions, pricing, and account support.

Set Response Limit

AI response limit per session caps how many times the bot replies during one conversation. Move the slider to the far right for unlimited responses.

Choose a finite limit when you need predictable conversation length or tighter usage control. Test the limit before deploying to customers.

Write Company Profile and Bot Purposes

This field tells the bot what your organization does and what it should handle. It accepts up to 1,000 characters. Include:

• Products or services the bot supports.
• Intended role, such as sales assistant or support agent.
• Topics within scope and topics it should decline.
• Important policies or service facts needed for common questions.
• Preferred response style and escalation guidance.

Use Help me write for a first draft, then verify every generated statement. Keep changing reference material, lengthy policies, and detailed catalogs in Bot Knowledge documents instead of overloading this field.

Save and Test

1. Check Bot Name, Bot Intro, and company profile for accuracy.
2. Confirm model and response limit.
3. Choose Save.
4. Test representative questions through a configured channel.
5. Confirm out-of-scope questions, response limits, and source-based answers behave as intended.