Skip to main content
After creating an AI Agent, the next step is configuration. Proper configuration ensures your AI responds accurately, uses the correct knowledge base, and performs actions automatically. To configure:
  1. Go to AI Agents
  2. Click on the AI you created
  3. You will see the following tabs:
    • Instructions
    • Actions
    • Knowledge
    • Settings

​
1. Instructions

Ai Instruction
The Instructions section defines how the AI should behave and respond to customers. The quality of your instructions directly impacts the quality of AI responses. Clear and structured instructions produce better results.

​
What You Can Add in Instructions

You can:
  • Define how the AI should talk (formal, friendly, sales tone, support tone)
  • Define what it should do
  • Add conditions
  • Insert actions
  • Attach files
  • Link WhatsApp flows (forms)
  • Apply contact tags

​
Using Actions in Instructions

Inside the Instructions box, type @ to insert an action. When you type @, a list of available actions will appear. You can select:
  • Assign conversation
  • Resolve conversation
  • Add a note
  • Tag a customer
  • Share a file
  • Trigger a WhatsApp flow
  • Custom API action
You can also create custom actions using API calls for advanced automation. This allows the AI to not just reply, but also perform backend operations automatically.

​
2. Knowledge

The Knowledge section allows you to attach relevant information sources to the AI Agent.
Ai Knowledge
To add knowledge:
  1. Click Manage Knowledge
  2. Select the knowledge base you want
  3. Attach it to the specific AI Agent
You can maintain multiple knowledge bases such as:
  • Support documents
  • Product information
  • FAQ documents
  • Policy documents
Example: If you have:
  • Support Knowledge Base
  • Sales Information Knowledge Base
You can attach the Support Knowledge only to the Support AI Agent. This ensures the AI responds only using relevant information.

​
3. Settings

The Settings tab allows you to control the AI model behavior and technical configuration.
Ai Settings

​
AI Provider

You can choose the AI provider:
  • OpenAI
  • Gemini
  • Anthropic
  • Others (based on availability)

​
Model Selection

You can select the model version such as:
  • gpt-5.2
  • gpt-5-mini
  • Other available models
Different models may vary in performance and cost.

​
Temperature

Temperature controls randomness in responses.
  • Lower values → More focused and deterministic responses
  • Higher values → More creative and varied responses
Example:
  • 0.1 → Very structured responses
  • 0.8 → More creative responses

​
Tokens

Defines the maximum number of tokens in the response. Higher token limit → Longer responses
Lower token limit → Shorter responses

​
Processing Capabilities

Ai Processing&response Settings
You can enable or disable processing for different media types. These toggles are ON by default.

​
Image Processing

If enabled, the AI can understand and respond to image inputs. You can turn OFF the toggle if not required.

​
Audio Processing

If enabled, the AI can process audio messages. You can disable it if your use case does not require audio handling.

​
Document Processing

If enabled, the AI can read and process document files shared by customers. You can turn it OFF if not required.

​
Response Settings

This section controls response behavior and delay settings.

​
Processing Message

Default message shown while AI is generating a response:
“Hold on a moment while I process your request…”
You can:
  • Edit this text
  • Turn OFF the toggle if you don’t want a processing message

​
Save Changes

After making all updates: Click Save Changes Your AI Agent configuration will be updated immediately. Your AI Agent is now fully configured and ready to automate conversations intelligently.