Agentic AI Prompt Guidelines
Learn how to write clear and effective instructions for your Agentic AI Agent’s capabilities, tasks, and tools.
What you’ll learn:
- How to properly configure global information for your Agentic AI Agent.
- How to write clear and structured instructions for your Agentic AI Agent’s capabilities, tasks, and interactions.
- Best practices for defining an Agentic AI Agent’s personality, knowledge base, and tools.
Global Information
To ensure your Agentic AI Agent functions effectively, define the following global details:
-
Company Name: The name of the business or organization the agent represents.
Company Name -
Company Background: A brief description of the company, including its industry, services, and target audience.
Company Background -
Agent Name: The designated name of your AI agent for a personalized experience.
Agent Name -
Welcome Message: The first message the agent sends to users when a conversation starts. It should be warm, engaging, and informative.
Welcome Message -
Agent Persona: A well-defined personality and communication style that aligns with the company’s branding.
Agent Persona -
Call Background: Contextual information dynamically fetched from real-time data sources to enhance interaction relevance.
-
Prior Knowledge: The agent’s pre-existing understanding of industry terms, processes, and frequently encountered scenarios.
Agent-Specific Information
Clearly outline the agent’s functional scope to ensure it performs efficiently.
Agent Capabilities
Define the agent’s primary functions and what it is trained to do. Examples include:
Agent Tasks
Task defines the way how the agent will interract with the caller. For example:
Agent Tools
List the tools, integrations, and APIs the agent will use to perform tasks efficiently.
Built-in Tools:
- Calendar Availability Check: To check available slots for scheduling.
- Calendar Booking: To book, reschedule, or cancel appointments.
- Send Email: For automated email responses and notifications.
- Transfer Call: To escalate conversations to human support.
Custom Tools:
- External REST API Integration: Users can add custom external REST APIs to extend the agent’s functionality.
Agent Knowledge Base
Specify the sources of information that the agent will rely on, including:
- Company FAQs: A curated list of frequently asked questions.
- Product/Service Documentation: Detailed guides on company offerings.
- Policy and Compliance Guidelines: Ensuring accurate responses aligned with company policies.
- Customer Interaction Logs: Past conversation data to improve response accuracy.
Best Practices
Clear Identification
Use clear and direct instructions to guide the agent’s behavior.
Tone & Personality
Maintain a consistent tone and personality aligned with the company’s branding.
User Clarity
Optimize prompts for user clarity and engagement.
Dynamic Adaptability
Ensure dynamic adaptability by leveraging real-time data sources.
Testing & Refinement
Test and refine the agent’s responses for accuracy and relevance.
By following these guidelines, you can build an effective, structured, and responsive Agentic AI Agent that enhances user experiences and business efficiency.