Skip to main content

​
AI Agents

AI Agents allow you to configure AI personas for automated responses and interactions.

​
Overview

AI Agents define how the AI responds to incoming messages. You can create specialized agents for different purposes like sales, support, or qualification. AI Agents Settings

​
Creating AI Agents

​
Agent Properties

PropertyDescription
NameAgent identifier (e.g., “Sales Closer”)
InstructionsPersona and behavior description
SharedMake visible to organization members
DefaultAuto-select in reply assistant

​
Defining Instructions

Write clear instructions including:
  • Persona: Who the agent is
  • Knowledge: What information it has access to
  • Behavior: How it should respond
  • Tone: Professional, casual, formal
  • Goals: What it should achieve
Example instructions: 
You are a friendly sales representative for a B2B SaaS company. 
Your goal is to qualify leads and schedule demo calls. 
Be professional but conversational. 
Ask qualifying questions to understand their needs. 
End by proposing a demo time if qualified.

​
Using AI Agents

​
Reply Assistant

AI Agents power the reply assistant:
  • Automatically suggests responses
  • Uses selected agent’s persona
  • Learns from conversation context

​
Campaign Responses

Configure AI agents for:
  • Auto-reply to campaign responses
  • Lead qualification flows
  • Meeting scheduling

​
Organization Support

​
Shared Agents

Agents can be shared with organization:
  • Visible to all organization members
  • Managed by creator
  • Can be unshared

​
Default Agent

Set a default agent:
  • Auto-selected in reply assistant
  • Can be overridden per conversation

​
API Endpoints

GET    /ai-agents/               - List agents
POST   /ai-agents/              - Create agent
PUT    /ai-agents/{id}          - Update agent
DELETE /ai-agents/{id}          - Delete agent

​
Best Practices

  1. Specific Personas: Create agents for specific use cases
  2. Clear Instructions: Be explicit about behavior
  3. Test Thoroughly: Test agents before production
  4. Iterate: Refine based on results
  5. Organization Use: Create shared agents for team consistency