Skip to main content
ChatSyncs left navigation with Chatbot Manager selected, the Integrations tab highlighted, the HTTP API sub-tab highlighted, and the HTTP API list with a Create button
Chatbot Manager → Integrations → HTTP API connects ChatSyncs to any external REST API (GET, POST, PUT, PATCH, DELETE) — the general-purpose way to hook a bot flow up to a system ChatSyncs doesn’t natively integrate with: a CRM, an order tracker, a payment gateway, a booking system, an ERP, or an inventory system.

​
Add HTTP API

Add HTTP API page showing the four-step flow: API Details, Request Data, Test and Verify, Response Mapping, with an AI API Builder panel on the right
Real-world example: an e-commerce store wants its bot to answer “Where is my order?” by calling its own Order API.
Customer: Where is my order? Bot: Please enter your Order ID. Customer: ORD12345
The bot sends an HTTP request to the store’s Order API and gets back: 
{
  "order_id": "ORD12345",
  "status": "Shipped",
  "courier": "BlueDart",
  "tracking_id": "BD987654321"
}
The bot replies:
Your order has been Shipped via BlueDart. Tracking ID: BD987654321
Here’s the full four-step setup that makes that possible:
1

API Details

Enter the API Name (e.g. Get Order Details), Method (e.g. POST), and API Endpoint URL (e.g. https://api.mystore.com/orders/details).
2

Request Data

Choose the request Body format — Default, Form Data, x-www-form-urlencoded, JSON, or Binary.Use a dynamic variable for whatever the customer typed:
{ "order_id": "{{order_id}}" }
If the customer enters ORD12345, the actual request sent becomes { "order_id": "ORD12345" }.Add whatever else the API needs:
  • Headers — e.g. Authorization: Bearer YOUR_ACCESS_TOKEN, Content-Type: application/json
  • Option Data — extra parameters some APIs require, e.g. language = en, store = bangalore
  • Cookies — only if the API needs session cookies for authentication; leave empty otherwise
3

Test & Verify

Enable Use Sample Data (Recommended) so detected variables get filled with sample values automatically, then click Send Test Request. A working API returns something like:
{
  "order_id": "ORD12345",
  "status": "Delivered",
  "courier": "BlueDart",
  "tracking_id": "BD987654321",
  "delivery_date": "2026-06-28"
}
Don’t move on until this test succeeds — an untested endpoint can fail silently mid-flow later.
4

Response Mapping

Map each field in the response to a variable the bot can use anywhere:
API ResponseVariable
status{{status}}
courier{{courier}}
tracking_id{{tracking_id}}
delivery_date{{delivery_date}}
The bot can now reply with any combination of these:
Your order is {{status}}.
Courier: {{courier}}
Tracking ID: {{tracking_id}}
Delivered On: {{delivery_date}}
which the customer sees as:
Your order is Delivered. Courier: BlueDart Tracking ID: BD987654321 Delivered On: 28 June 2026

​
AI API Builder (Beta)

Instead of manually filling in every field, the AI API Builder panel next to the form can generate the whole configuration for you from one of:
  • Describe in English — e.g. “I have an Order API that accepts an Order ID and returns the order status, courier name, tracking ID, and delivery date.”
  • Paste cURL
  • Paste your API Docs
  • Paste JSON
The AI fills in the API Endpoint, Request Method, Headers, Request Body, Variables, and Response Mapping automatically.
Always review and Send Test Request on an AI-generated configuration before relying on it in a live bot flow — treat it as a first draft, not a finished, verified integration.

​
Complete workflow

1

Enter API Details

API Name, Method, and Endpoint URL.
2

Configure the request

Body format (JSON, Form Data, etc.), Headers, Option Data, Cookies.
3

Send a test request

Use sample data, then confirm the response looks right.
4

Map the response

Assign each response field to a bot variable.
5

Save the API

It’s now available to any flow.
6

Use the mapped variables

Reference them anywhere in your chatbot flow.

​
Use cases

  • CRM integration — look up or update a customer record.
  • Order tracking — the walkthrough above.
  • Payment verification — confirm a transaction before continuing a flow.
  • Customer lookup — pull account details by phone number or email.
  • Booking systems — check availability or confirm a reservation.
  • ERP / inventory integration — check stock before promising a delivery date.

​
Frequently asked

Use Chatbot Manager → Integrations → HTTP API → Create: enter the endpoint and method, configure the request body with a dynamic variable like {{order_id}}, test it, then map the response fields (status, courier, tracking ID, etc.) to bot variables you can use in a reply.
Yes — use the AI API Builder (Beta) panel: describe the API in plain English, paste a cURL command, paste its docs, or paste raw JSON, and it generates the endpoint, headers, body, variables, and response mapping for you. Always test the result before using it live.
Add it under Headers in the Request Data step — most APIs expect something like Authorization: Bearer YOUR_ACCESS_TOKEN. If the API instead uses session cookies, use the Cookies section.
Test an HTTP API before relying on it in a live bot flow — use Send Test Request with sample data first. An untested endpoint can silently fail mid-conversation, leaving the customer with a broken reply instead of an error you can catch early.