AI Persona
Overview
The AI Persona defines how a Conversational Agent presents itself and behaves during conversations.
A well-designed persona helps create consistent, predictable, and engaging interactions by defining both the agent's visual identity and conversational behavior.
The AI Persona influences:
- Agent identity
- Communication style
- Personality
- Response behavior
- User experience
- Brand representation
Persona Components
In IB-X, the AI Persona consists of two primary areas:
Visual Identity
Visual Identity controls how the agent is presented to users.
Typical settings include:
- Agent Name
- Agent Role
- Avatar
- Theme Color
These settings help users recognize and interact with the agent consistently across supported channels.
Behavioral Identity
Behavioral Identity defines how the agent thinks, responds, and interacts during conversations.
Behavioral characteristics are primarily controlled through the instructions configured in the Conversational Agent activity.
These instructions define:
- Responsibilities
- Communication style
- Response formatting
- Business rules
- Tool usage behavior
- Escalation guidance
- Compliance requirements
- Safety restrictions
Together, these instructions form the operational personality of the agent.
Agent Name
The Agent Name represents the display name shown to users during conversations.
Examples:
Lily
IT Support Assistant
HR Assistant
Customer Care Agent
The selected name should clearly communicate the purpose of the agent.
Agent Role
The Agent Role provides additional context regarding the responsibility of the agent.
Examples:
Conversational Agent
Support Specialist
Knowledge Assistant
Sales Advisor
The role helps users understand the capabilities and purpose of the agent.
Avatar
The Avatar represents the visual identity of the agent.
The selected avatar is displayed in supported conversational experiences and helps create a more engaging user experience.
Organizations often choose avatars that align with their branding and intended audience.
Theme Color
The Theme Color controls the primary visual color used throughout the conversational interface.
This color is typically applied to:
- Chat headers
- Action buttons
- Highlighted elements
- Visual accents
Organizations often align the theme color with corporate branding guidelines.
Defining Agent Behavior
While visual settings define how the agent appears, the instructions configured in the Conversational Agent activity define how the agent behaves.
Examples of behavioral instructions include:
Support Agent
You are a customer support assistant.
Provide concise and professional responses.
Use enterprise knowledge when available.
Escalate issues when sufficient information is unavailable.
IT Helpdesk Agent
You are an IT support assistant.
Help users troubleshoot common technical issues.
Provide step-by-step guidance.
Do not make assumptions about system configuration.
Knowledge Assistant
You are a knowledge assistant.
Answer questions only using available enterprise knowledge.
If information cannot be found, clearly state that the information is unavailable.
Persona Design Guidelines
When designing a persona:
Define a Clear Purpose
Users should immediately understand the role of the agent.
Avoid creating agents with overly broad responsibilities.
Use Consistent Communication
Define:
- Formal or informal tone
- Response length
- Terminology preferences
- Formatting standards
Consistency improves user trust and predictability.
Establish Boundaries
Clearly define:
- Supported topics
- Unsupported topics
- Escalation requirements
- Tool usage restrictions
Boundaries help reduce unintended behavior.
Align with Organizational Branding
The visual appearance, tone, and behavior of the agent should align with organizational standards and customer expectations.
Best Practices
- Give the agent a meaningful name.
- Use a role that clearly communicates purpose.
- Choose avatars that align with organizational branding.
- Keep instructions focused and specific.
- Define clear behavioral boundaries.
- Test the persona using realistic conversation scenarios.
- Continuously refine the persona based on user feedback.
Related
- Building a Conversational Agent
- Conversational Agent Activity
- Knowledge Grounding
- Tools and Actions