Skip to main content

Create and test your first AI agent

Configure an AI agent and test an example order.

This narrated walkthrough demonstrates the same workflow documented in the written, step-by-step guide directly below.

Create an Ordering.co AI agent​

This guide walks through creating an Ordering.co AI agent, configuring its widget, and testing an example order from start to finish.

1. Bringing your agent to life​

The journey starts in the AI Agents dashboard. Think of this as your "mission control," where you can monitor every bot in your fleet at a glance.

  • Create Your Agent: Simply click Create Agent to begin.
  • The setup: Select Finitless, the provider name shown in this flow, and choose the Chat Provider type.
  • Brand Integration: This is the fun part—choosing your franchise! For this example, we’re selecting Pizza Pro’s.
  • Go Live: Once you hit create, your agent is officially born and ready for work.

2. Branding and customization​

After creating the agent, review its default site URL and add a custom domain if the account and domain configuration support one.

Designing the widget​

The Widget Configuration is where you decide how your AI introduces itself to the world.

  • Floating Chat: A sleek, WhatsApp-style bubble that stays in the corner of the screen.
  • Popup Modal: A centered window that really grabs the user’s attention.
  • The Look: You have full control over the colors. You can choose primary and secondary gradients that match your brand identity.
  • The Feel: You can even set the button size and position to ensure the user experience is as smooth as possible.

3. The customer experience (in action)​

Let’s step into the customer’s shoes. We head over to the live site and tell the AI, “I’m hungry”.

  • Secure Entry: To keep things secure, the system asks for an email. Once the customer enters the six-digit validation code, they are logged in and ready to roll.
  • Personalization: The AI is smart! It recognizes past orders and asks if we want delivery or pickup.
  • Ease of Use: If we choose delivery, it already knows our usual address on Fifth Ave.
  • The Menu: The agent confirms the details and presents the main categories: pizzas, drinks, and desserts.

Ordering is a breeze​

In the recorded example, the customer requests "I want a large pepperoni pizza with original crust, no extra toppings, and a Coke". The agent confirms the cart and offers a dessert.

4. The final review and checkout​

At the checkout, the agent summarizes the entire order clearly.

  • Payment Preferences: The AI asks how the customer would like to pay.
  • Special Requests: If the customer chooses cash and mentions they need change for a hundred, the example shows that note in the order.
  • Clear Communication: The example also reminds the customer to confirm the cash payment with the driver upon arrival.

5. Behind the scenes: management​

The Ordering.co Admin Panel in this example shows the resulting order details.

  • Real-Time Tracking: Under the Orders tab, you can see every new request in real-time.
  • Deep Dive: You can click into the details of any order to see the customer information, the specific items requested, and the total amount due.

Tips, tricks & best practices​

  • Test Your Gradients: Use high-contrast colors for your chat widget so it doesn't get lost in your website's background.
  • Keep Menus Updated: Confirm the menu data available to the agent before testing an order.
  • Analyze the Logs: Review order details for repeated requests that are not represented in the menu.
  • Review Mobile Placement: Confirm that the button remains visible and reachable in the mobile preview.

Conclusion​

The test confirms how the configured agent presents a menu, records order details, and hands the order to the dashboard for review.