Skip to main content

Requirements

Env variables

Customer service agent

Taken from OpenAI Agents Example Folder. Flow:
  1. User Input: The user asks a question or makes a request.
  2. Triage Agent: This agent first handles the input, figuring out what the user needs and deciding which specialized agent should take over.
  3. Specialized Agents: Based on the request, the Triage Agent may pass the conversation to either the FAQ Agent or the Seat Booking Agent.
    • FAQ Agent: Answers common questions using a predefined tool.
    • Seat Booking Agent: Manages seat booking requests and updates seat information.
  4. Context Update: The system updates the context (like flight number, seat number, etc.) based on the user’s input and the actions taken by the agents.
  5. Logging/Tracing: All interactions and updates are logged and traced using the Maxim SDK for monitoring and analysis.
  6. Output: The final response is generated and sent back to the user.
This system ensures that user requests are efficiently routed to the right agent, and all interactions are tracked for quality and performance monitoring.

Initializing Maxim SDK

  • Maxim SDK picks up MAXIM_API_KEY and MAXIM_LOG_REPO_ID from environment variables.
  • You can pass them as parameters if you would like to.
  • Learn more here.

Run agent

Maxim dashboard

You can view the trace of the agents’ interactions on the Maxim dashboard, which provides detailed insights and visualizations of the entire process. Openai Agents Customer Support Gi