So far I've discussed what AI Agents are and how they will be able to assist and connect with customers.

Imagine for a moment that you are the owner of WindowsRUs, a small business that manufactures and installs Windows.

• A customer calls in to say that a seal in one of the installed windows is broken.
• The window is still under warranty and the customer would like to schedule a technician to come out and replace it.
• How would an AI Agent resolve this issue?

It’s a trick question because you wouldn’t be able to resolve it, today, with just one AI Agent (AIA).  You would need several AIAs working in collaboration to resolve this issue.

Before sharing a collaborative scenario, it's worth taking a moment to understand how AI Agents communicate or talk to each other. Here are 3 ways:

1. Data Sharing: The agents share information using a central database (e.g., CRM, ERP, etc.) or other data resources (e.g., documents or spreadsheets)
2. API Calls: Think of API calls as like sending a text message or email. When one agent needs information from another, it sends a request. The receiving agent then responds with the requested data.
3. Natural Language Processing (NLP): On the front end of customer service, AIAs interact directly with customers either via text or spoken words.

Workflow Summary:So, when a customer calls in, the 'conversation AI' uses NLP to understand the issue. It then sends a request to the scheduling AI, which uses its database to find available times. The scheduling AI sends the appointment details to the technician routing AI, which uses GPS data and traffic information to plan the best route. And so on.

Case Study: Customer Service for WindowsRUs

Back to my opening example. A collaborative network of AI agents would be deployed to handle this customer inquiry efficiently. Here's how they might work together:

1. Customer Service AI: A chatbot designed to understand and respond to customer inquiries would handle the initial interaction. This chatbot would gather basic information, such as the customer's name, contact details, and the specific window in question. It would also verify the warranty status.

2. Scheduling AI: Once the chatbot has collected the necessary information, it will transfer the inquiry to a scheduling AI. This AI would consult the company's calendar and technician availability to find the earliest suitable appointment time. It would also factor in the customer's preferred time and location.

3. Technician Routing AI: After a suitable appointment is scheduled, a technician routing AI will be activated. This AI would determine the optimal route for the technician based on their location, the customer's address, and traffic conditions. It would also provide real-time updates to the technician's GPS device.

4. Inventory Management AI: To ensure that the necessary replacement parts are available for the technician's visit, an inventory management AI will be consulted. This AI would check the company's inventory records to determine if the specific window seal is in stock. If not, it would initiate an order to replenish the supply.

5. Customer Follow-up AI: After the technician has completed the repair or replacement, a customer follow-up AI would reach out to the customer to ensure their satisfaction. This AI would collect feedback on the technician's service, the quality of the replacement window, and the overall experience.

By working together, these AI agents can streamline the customer service process, reduce response times, and improve overall customer satisfaction.

The future of AI is collaborative whether agent-to-agent-to-agent or  agent-to-agent-to-agent with a Human-In-The-Loop (HITL).