Best Practices

Agent Naming Conventions in raiaAI Launch Pad

Proper agent naming is crucial for organization and team collaboration, especially when managing multiple agents for different clients. Follow these naming conventions to ensure clarity and easy identification:

Function-First Naming: Always start agent names with their primary function, making it immediately clear what the agent does. This approach scales well as you build multiple agents for the same client.

Recommended Naming Pattern: [Function] Agent - [Client/Department] - [Specific Use Case if needed]

Good Examples:

• "Support Agent - TechCorp"

• "Sales Agent - RetailPlus - Lead Qualification"

• "Analyst Agent - FinanceInc - Report Generation"

• "Support Agent - MedDevice - Technical Troubleshooting"

Poor Examples to Avoid:

• "Mary" (no indication of function)

• "TechCorp Bot" (unclear what it does)

• "Agent 1" (not descriptive)

• "Customer Helper" (too vague)

Benefits of Function-First Naming:

• Immediate clarity on agent purpose when browsing the Launch Pad

• Easy sorting and organization by function type

• Simplified handoffs between team members

• Clear audit trails and reporting

• Easier troubleshooting and maintenance

Model Selection Standards

raiaLabs primarily uses OpenAI's most advanced models to ensure optimal performance and capabilities for client agents:

Primary Models:

• GPT-4o: The standard model for most agent deployments, offering excellent performance across all agent types

• GPT-5: Used for the most demanding applications requiring cutting-edge capabilities

Model Selection Guidelines:

• Support Agents: GPT-4o provides excellent knowledge retrieval and conversational abilities

• Sales Agents: GPT-4o or GPT-5 for sophisticated persuasion and objection handling

• Analyst Agents: GPT-4o for most analytical tasks, GPT-5 for complex data analysis requiring advanced reasoning

Consistency Across Client Deployments: Using standardized models ensures consistent performance and makes it easier to:

• Predict agent behavior and capabilities

• Troubleshoot issues across different client implementations

• Maintain training materials and best practices

• Scale successful patterns to new clients

This foundational understanding of agent types, naming conventions, and model standards will inform all subsequent training modules, helping you make appropriate decisions about instruction design, knowledge base development, and workflow integration based on the specific agent type you're building.