For businesses that handle digital products like airtime reselling, ticket bookings, subscription management, or any service with a backend API, support often requires more than just answering frequently asked questions. Customers need real-time help with their orders, payments, and accounts—and they want it now, without waiting for a support agent to become available.
This is where our Custom API integration feature shines. Instead of just providing information, your AI support agent can actually take action on behalf of your customers, connecting directly to your existing systems to resolve issues instantly.
What You Can Do With Custom API Integration
The possibilities are almost endless, but here are some of the most powerful use cases our customers are already implementing:
- Check if a customer's payment was successful and automatically send airtime or digital goods
- Verify order status and provide real-time updates without any agent intervention
- Reset passwords, extend trials, or adjust subscription details directly through the chat
- Process refunds or credit accounts based on your existing business logic and policies
- Check inventory levels and confirm product availability before a customer makes a purchase
- Validate coupon codes and apply discounts automatically during conversations
- Retrieve and share invoices, receipts, or other documents securely
How to Set Up Your Custom API: Step by Step
Getting started with Custom API integration is straightforward. Follow these steps to connect your first API endpoint:
1. Prepare Your API Endpoint
First, make sure your API is accessible from the public internet (or use a tunneling tool like ngrok for testing). Your API should accept JSON requests and return JSON responses. You’ll need to know the URL, HTTP method (GET, POST, PUT, etc.), and any required headers or authentication details.
2. Add the API in Your Dashboard
Navigate to the 'Custom APIs' section in your dashboard and click 'Add New API'. Fill in the details: name (for your reference), endpoint URL, HTTP method, and any headers required for authentication (like API keys, Authorization tokens, etc.).
3. Define Data Collection
Next, specify what information the AI needs to collect from the customer before calling the API. For example, if you’re verifying payments, you might need a transaction ID or phone number. If you’re checking order status, you’ll need an order number.
4. Set Up Response Handling
Define how the AI should interpret the API response. What should it say if the request was successful? What if there was an error? You can set up different message templates for different response scenarios to ensure clarity for your customers.
5. Test and Launch
Before launching to all customers, test the integration thoroughly in a staging environment. Send test queries to the AI to make sure it collects the right information, calls the API correctly, and handles responses appropriately.
Airtime Reseller Example: How It Works in Practice
Let’s walk through a real-world example of how this works for an airtime reselling business:
The Customer’s Query
A customer messages in: 'Hi, I just paid N500 for airtime to 08012345678 but I haven’t received it yet. Can you help?'
The AI Responds and Collects Information
The AI recognizes this as a payment/airtime issue and asks: 'I’m sorry to hear that! To help, could you please share your transaction ID?' The customer provides the ID: 'TX123456789'
The AI Calls Your Payment Verification API
First, the AI calls your payment verification API with the transaction ID. The API confirms that the payment was successful and provides details about the transaction.
The AI Calls Your Airtime Delivery API
Next, the AI calls your airtime delivery API with the phone number and amount. The delivery API processes the request and confirms that the airtime has been sent successfully.
The AI Confirms With the Customer
Finally, the AI sends a message to the customer: 'Great news! I’ve confirmed your payment and sent your airtime to 08012345678. You should receive it any moment now. Is there anything else I can help you with?'
The entire process takes 10-15 seconds, no support ticket is created, and your team can focus on more complex issues. That’s the power of automation!
Security Best Practices for Custom APIs
When connecting your business systems, security is paramount. Here are some best practices to keep in mind:
- Always use HTTPS for your API endpoints to encrypt data in transit
- Use API keys or OAuth tokens instead of hardcoding credentials
- Limit API permissions to only what the AI needs (don’t use an admin token)
- Implement rate limiting to prevent abuse
- Log API calls for auditing and debugging purposes
- Never store sensitive customer data longer than necessary
Ready to automate your most repetitive support tasks? Start by adding your first custom API in the dashboard today. If you need help, our support team is here to guide you through the process.