When you’re thinking about building an agent, it’s best to imagine it as a specialized expert focused on a specific task. Instead of creating one agent to do everything, think about breaking down complex processes into smaller, more focused agents.

​
Think modular

Consider this example from the flight booking business, where instead of a single agent handling everything, we break it down into specialized agents:

Ticket cancellation agent

Handles only flight cancellations efficiently.

New ticket booking agent

Dedicated to just booking new flights, simplifying the booking process.

Flight rescheduling agent

Specifically designed to manage flight rescheduling requests.
This focused approach makes each agent more effective, easier to manage, and leads to a better user experience. It also allows for easier optimization and improvement of each agent individually.

​
Identifying agent opportunities

A great way to find opportunities for AI agents is to consider where conversations are already happening naturally.

​
Example use case: Flight onboarding assistant

Imagine customers frequently call about baggage limits or pet policies. A flight onboarding assistant agent can handle these common inquiries, automating tasks like:

Flight tracking

Providing up-to-the-minute flight status updates.

Baggage information

Detailing baggage allowances and excess baggage policies.

Policy explanations

Clarifying pet policies and other flight-related guidelines.

​
Benefits

​
Example Use Cases