External AI Agent Support — Implementation Guide

Prerequisites

A WhatsApp number already connected to ChatSyncs.
An external server or workflow tool reachable from the internet, with your own AI agent already built on it — an n8n workflow with an OpenAI Chat Model and Simple Memory node, in this build. Building and training that AI agent is entirely outside ChatSyncs; this guide only covers the WhatsApp connection.
Your ChatSyncs API token and Phone Number ID — see Finding IDs — needed to call the Developer API from your own server.
Read Bot Settings first for the full Webhook tab reference; this guide only covers the Incoming Webhook toggle specifically.

Implementation Guide

Step 1 — Enable the Incoming Webhook

Objective: Have ChatSyncs forward every customer message to your AI agent’s server the moment it arrives. Explanation: The Incoming Webhook is a toggle on the bot’s Webhook settings tab — once enabled, every message a customer sends is POSTed to the URL you enter, in addition to whatever ChatSyncs’ own bot flow would otherwise do with it. Actions to perform:

Open Bot Settings → Webhook.
Turn on Trigger Webhook for Incoming Message.
Enter your AI agent’s webhook URL (its n8n Webhook node’s URL, in this build).
Click Publish Changes.

ChatSyncs Bot Settings Webhook tab with Trigger Webhook for Incoming Message enabled and the webhook URL set to an n8n webhook endpoint, with the Bot Settings button and Publish Changes button highlighted

Expected result: Every new customer message now also arrives at your AI agent’s server as a POST request, alongside whatever ChatSyncs itself does with it.

Step 2 — Your AI agent generates a reply

Objective: Let your existing AI agent decide what to say, using its own tools and memory. Explanation: This step runs entirely outside ChatSyncs, on your own infrastructure. In this build: an n8n Webhook node (“incoming Messages”) receives the POST from Step 1, a Set node (“data”) shapes the payload, an AI Agent node (backed by an OpenAI Chat Model and a Simple Memory node, with room for Tools) reads the message and produces a reply, an output Split node breaks that reply into its message-type parts, and a Switch node then routes each part down one of three branches — interactive buttons, plain text, or an image with interactive buttons.

n8n workflow canvas showing an incoming Messages webhook node feeding a data Set node into an AI Agent node backed by an OpenAI Chat Model and Simple Memory, through an output Split node into a Switch node that branches to Interactive, Text, and image and interactive nodes, each POSTing to the ChatSyncs platform API

Actions to perform (on your own server):

incoming Messages (Webhook) node: receives the incoming-message payload from Step 1.
data (Set) node: pulls the fields the AI Agent needs out of that payload.
AI Agent node: generates the reply, with whatever chat model, memory, and tools your setup already uses.
output Split node: splits the AI’s structured output into its component parts (text, buttons, image).
Switch node: routes each part to the matching branch — interactive buttons, plain text, or image-plus-buttons — based on the message type the AI Agent returned.

Expected result: A finished reply, ready to send, plus a decision about which message format (or combination of formats) to use for it. Notes: None of this — the model, the memory, the tools, the split/switch routing — is a ChatSyncs feature. ChatSyncs never sees this part of the workflow; it only sees what comes back in Step 3.

Step 3 — Send the reply back through the ChatSyncs API

Objective: Deliver the AI’s reply to the customer on WhatsApp. Explanation: ChatSyncs’ Send Text Message and Send Interactive Buttons API endpoints let any external server push a message straight to a customer’s WhatsApp chat. In this build, each of the Switch node’s three branches ends by POSTing straight to the ChatSyncs platform API — one for interactive buttons, one for plain text, and one for an image sent together with interactive buttons. Actions to perform (on your own server):

Text branch: POST /whatsapp/send with apiToken, phone_number_id, the customer’s phone_number, and message set to the AI’s reply text.
Interactive branch: POST /whatsapp/send/interactive-buttons with the same apiToken, phone_number_id, and phone_number, plus message for the body text and buttons as a JSON array of up to 3 options.
Image + interactive branch: the same interactive-buttons call, with an added image URL so the buttons are delivered alongside a picture instead of body text alone.

Expected result: The customer receives the AI’s answer on WhatsApp — as a plain message, tappable Reply Buttons, or an image with Reply Buttons — within the same conversation they messaged into. Notes: These endpoints send session messages — only deliverable within 24 hours of the customer’s last message. Outside that window, a Message Template is required instead; see Send Text Message for that limit.

Copy & test this workflow

The exact n8n workflow used throughout Step 2 and Step 3 above — the incoming webhook, the AI Agent with its Chat Model and Memory, the output split, and all three send branches (interactive, text, image + interactive) — is available below as a real exported workflow JSON:

View the full n8n workflow JSON

Select all and copy the block below, then paste it (Ctrl+V / Cmd+V) directly onto your n8n canvas — n8n imports pasted workflow JSON automatically, without needing to save or upload a file first.

{
  "name": "Ai Agent",
  "nodes": [
    {
      "parameters": {
        "promptType": "define",
        "text": "={{ $('incoming Messages').item.json.body.user_message }}",
        "options": {
          "systemMessage": "=You are \"Sunny\", the WhatsApp AI Assistant for Sunrise Family Clinic.\n\nYour job is to receive ONE WhatsApp message at a time and return EXACTLY ONE valid JSON object.\n\n==========================\nOUTPUT RULES\n==========================\n\n- Return ONLY valid JSON.\n- Never return markdown.\n- Never return explanations.\n- Never return code blocks.\n- Never return text outside JSON.\n- Always return exactly ONE JSON object.\n\n==========================\nMESSAGE TYPES\n==========================\n\nThere are three response formats.\n\n1. Text\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"Your response\"\n}\n\n2. Button\n\n{\n  \"message_type\": \"button\",\n  \"body\": \"Message\",\n  \"footer\": \"Sunrise Family Clinic\",\n  \"buttons\": [\n    {\n      \"id\": \"button_id\",\n      \"title\": \"Button Title\"\n    }\n  ]\n}\n\n3. Image\n\n{\n  \"message_type\": \"image\",\n  \"image_url\": \"https://example.com/image.jpg\",\n  \"footer\": \"Sunrise Family Clinic\",\n  \"buttons\": [\n    {\n      \"id\": \"button_1\",\n      \"title\": \"Book Appointment\"\n    },\n    {\n      \"id\": \"button_2\",\n      \"title\": \"Contact Clinic\"\n    },\n    {\n      \"id\": \"button_3\",\n      \"title\": \"View Services\"\n    }\n  ]\n}\n==========================\nBUTTON RULES\n==========================\n\nUse button responses ONLY when presenting choices to the user.\n\nButton Rules\n\n- Maximum 3 buttons.\n- Button title must be 20 characters or fewer.\n- Button IDs must use lowercase and underscores.\n- Every button ID must be unique.\n- Never create more than 3 buttons.\n- Never mix text and buttons in one response.\n- Footer must always be:\n  \"Sunrise Family Clinic\"\n\n==========================\nWelcome menu\n==========================\n\nIf the user says:\n\nhi\nhello\nhey\nhii\ngood morning\ngood afternoon\ngood evening\nstart\nmenu\nhelp\n\nReturn\n\n{\n  \"message_type\": \"image\",\n  \"image_url\": \"https://your-domain.com/welcome-image.jpg\",\n  \"body\": \"👋 Welcome to Sunrise Family Clinic!\\n\\nHow can we help you today?\",\n  \"footer\": \"Sunrise Family Clinic\",\n  \"buttons\": [\n    {\n      \"id\": \"book_appointment\",\n      \"title\": \"📅 Book\"\n    },\n    {\n      \"id\": \"services\",\n      \"title\": \"🩺 Services\"\n    },\n    {\n      \"id\": \"contact_us\",\n      \"title\": \"📞 Contact\"\n    }\n  ]\n}\n==========================\nBUTTON NAVIGATION\n==========================\n\nWhen the user clicks a button ID, treat it as a menu selection.\n\nEach button may return another button menu OR a text response depending on the situation.\n\n----------------------------------\nBOOK APPOINTMENT\n----------------------------------\n\nIf user sends\n\nbook_appointment\n\nReturn\n\n{\n  \"message_type\": \"button\",\n  \"body\": \"Please choose the department.\",\n  \"footer\": \"Sunrise Family Clinic\",\n  \"buttons\": [\n    {\n      \"id\": \"general_physician\",\n      \"title\": \"👨‍⚕️ General\"\n    },\n    {\n      \"id\": \"dental\",\n      \"title\": \"🦷 Dental\"\n    },\n    {\n      \"id\": \"more_services\",\n      \"title\": \"➡️ More\"\n    }\n  ]\n}\n\n----------------------------------\nMORE SERVICES\n----------------------------------\n\nIf user sends\n\nmore_services\n\nReturn\n\n{\n  \"message_type\": \"button\",\n  \"body\": \"Choose a department.\",\n  \"footer\": \"Sunrise Family Clinic\",\n  \"buttons\": [\n    {\n      \"id\": \"dermatology\",\n      \"title\": \"🧴 Skin\"\n    },\n    {\n      \"id\": \"physiotherapy\",\n      \"title\": \"🏃 Physio\"\n    },\n    {\n      \"id\": \"health_checkup\",\n      \"title\": \"🩺 Checkup\"\n    }\n  ]\n}\n\n----------------------------------\nSERVICES\n----------------------------------\n\nIf user sends\n\nservices\n\nReturn\n\n{\n  \"message_type\": \"button\",\n  \"body\": \"Which service would you like to know about?\",\n  \"footer\": \"Sunrise Family Clinic\",\n  \"buttons\": [\n    {\n      \"id\": \"consultation_info\",\n      \"title\": \"👨‍⚕️ Consult\"\n    },\n    {\n      \"id\": \"health_packages\",\n      \"title\": \"📋 Packages\"\n    },\n    {\n      \"id\": \"working_hours\",\n      \"title\": \"🕘 Timings\"\n    }\n  ]\n}\n\n----------------------------------\nCONTACT\n----------------------------------\n\nIf user sends\n\ncontact_us\n\nReturn\n\n{\n  \"message_type\": \"button\",\n  \"body\": \"How would you like to contact us?\",\n  \"footer\": \"Sunrise Family Clinic\",\n  \"buttons\": [\n    {\n      \"id\": \"clinic_phone\",\n      \"title\": \"📞 Call\"\n    },\n    {\n      \"id\": \"clinic_address\",\n      \"title\": \"📍 Address\"\n    },\n    {\n      \"id\": \"clinic_email\",\n      \"title\": \"✉️ Email\"\n    }\n  ]\n}\n\n==========================\nLEAF BUTTONS\n==========================\n\nWhen the user selects a final button, return TEXT only.\n\nExample\n\ngeneral_physician\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"Sure! Please share your preferred date, time, patient name, and contact number to book a General Physician consultation.\"\n}\n\nExample\n\ndental\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"Please share your preferred date, time, patient name, and contact number to schedule a Dental appointment.\"\n}\n\nExample\n\ndermatology\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"Please share your preferred date, time, patient name, and contact number to schedule a Dermatology consultation.\"\n}\n\nExample\n\nphysiotherapy\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"Please share your preferred date, time, patient name, and contact number to schedule a Physiotherapy session.\"\n}\n\nExample\n\nhealth_checkup\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"Please let us know your preferred date and contact number. We'll help you book a Health Checkup.\"\n}\n\nExample\n\nconsultation_info\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"We provide consultations for General Medicine, Dental Care, Dermatology, Physiotherapy, and Health Checkups.\"\n}\n\nExample\n\nhealth_packages\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"We offer preventive health checkup packages for individuals and families. Please let us know your age group or requirements.\"\n}\n\nExample\n\nworking_hours\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"Our clinic is open Monday to Saturday from 9:00 AM to 8:00 PM.\"\n}\n\nExample\n\nclinic_phone\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"You can reach Sunrise Family Clinic at +91-9876543210 during working hours.\"\n}\n\nExample\n\nclinic_address\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"Sunrise Family Clinic, Main Road, Bangalore.\"\n}\n\nExample\n\nclinic_email\n\n{\n  \"message_type\": \"text\",\n  \"text\": \"You can email us at contact@sunriseclinic.com.\"\n}\n\n==========================\nNORMAL CONVERSATION\n==========================\n\nFor every other user message:\n\n- Understand the intent.\n- Answer professionally.\n- Keep responses short.\n- Ask for missing information if required.\n- Use TEXT format only.\n\n==========================\nCLINIC INFORMATION\n==========================\n\nClinic Name:\nSunrise Family Clinic\n\nServices\n\n- General Physician Consultation\n- Dental Care\n- Dermatology\n- Physiotherapy\n- Health Checkups\n\nWorking Hours\n\nMonday to Saturday\n\n9:00 AM to 8:00 PM\n\n==========================\nFINAL VALIDATION\n==========================\n\nBefore returning the response, verify:\n\n✓ Valid JSON\n✓ Only ONE JSON object\n✓ Greeting → Welcome buttons\n✓ Menu selections → Appropriate submenu or text\n✓ Normal conversation → Text only\n✓ Maximum 3 buttons\n✓ Button titles ≤20 characters\n✓ No markdown\n✓ No explanations\n✓ No extra text\n\nReturn ONLY the JSON object."
        }
      },
      "type": "@n8n/n8n-nodes-langchain.agent",
      "typeVersion": 3.1,
      "position": [-96, 16],
      "id": "6c18c6a1-9eee-4d7f-a780-cbf39689cdcf",
      "name": "AI Agent"
    },
    {
      "parameters": {
        "sessionIdType": "customKey",
        "sessionKey": "={{ $('incoming Messages').item.json.body.chat_id }}"
      },
      "type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
      "typeVersion": 1.4,
      "position": [-48, 288],
      "id": "2204a9f6-8716-4bde-80d4-235922ce6603",
      "name": "Simple Memory"
    },
    {
      "parameters": {
        "model": { "__rl": true, "mode": "list", "value": "gpt-5-mini" },
        "builtInTools": {},
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
      "typeVersion": 1.3,
      "position": [-208, 272],
      "id": "c597e179-bf47-4d55-bd5b-2bbdc1250383",
      "name": "OpenAI Chat Model",
      "credentials": {
        "openAiApi": { "id": "48jg0IN7QfVXoF59", "name": "OpenAi Freelance" }
      }
    },
    {
      "parameters": { "public": true, "options": {} },
      "type": "@n8n/n8n-nodes-langchain.chatTrigger",
      "typeVersion": 1.4,
      "position": [-288, -416],
      "id": "6712cbac-076b-434a-8b3b-d29ab3944e62",
      "name": "When chat message received",
      "webhookId": "51b26266-48be-4cf5-a9f5-10d9053e809c"
    },
    {
      "parameters": {
        "method": "POST",
        "url": "https://platform.chatsyncs.com/api/v1/whatsapp/send/interactive-buttons",
        "sendBody": true,
        "contentType": "form-urlencoded",
        "bodyParameters": {
          "parameters": [
            { "name": "apiToken", "value": "={{ $('data').item.json.api_key }}" },
            { "name": "phone_number_id", "value": "={{ $('data').item.json.phonenumber_id }}" },
            { "name": "phone_number", "value": "={{ $('incoming Messages').item.json.body.chat_id }}" },
            { "name": "message", "value": "={{ $json.body }}" },
            { "name": "buttons", "value": "={{ $json.buttons }}" },
            { "name": "button_footer_text", "value": "={{ $json.footer }}" }
          ]
        },
        "options": {}
      },
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.4,
      "position": [880, -240],
      "id": "f2902d11-d14e-4f75-9984-31f8164c51ac",
      "name": "Interactive"
    },
    {
      "parameters": {
        "rules": {
          "values": [
            {
              "conditions": {
                "options": { "caseSensitive": true, "leftValue": "", "typeValidation": "strict", "version": 3 },
                "conditions": [
                  {
                    "leftValue": "={{ $json.message_type }}",
                    "rightValue": "=button",
                    "operator": { "type": "string", "operation": "equals" },
                    "id": "53efed05-ee28-48c6-a092-9e4819106630"
                  }
                ],
                "combinator": "and"
              },
              "renameOutput": true,
              "outputKey": "Interactive"
            },
            {
              "conditions": {
                "options": { "caseSensitive": true, "leftValue": "", "typeValidation": "strict", "version": 3 },
                "conditions": [
                  {
                    "id": "756985a8-5e52-4759-8aa2-7ff125cac7ac",
                    "leftValue": "={{ $json.message_type }}",
                    "rightValue": "text",
                    "operator": { "type": "string", "operation": "equals" }
                  }
                ],
                "combinator": "and"
              },
              "renameOutput": true,
              "outputKey": "text"
            },
            {
              "conditions": {
                "options": { "caseSensitive": true, "leftValue": "", "typeValidation": "strict", "version": 3 },
                "conditions": [
                  {
                    "id": "5939537f-f5b4-4a08-8231-2910562eb6be",
                    "leftValue": "={{ $json.message_type }}",
                    "rightValue": "image",
                    "operator": { "type": "string", "operation": "equals", "name": "filter.operator.equals" }
                  }
                ],
                "combinator": "and"
              },
              "renameOutput": true,
              "outputKey": "Image"
            }
          ]
        },
        "options": {}
      },
      "type": "n8n-nodes-base.switch",
      "typeVersion": 3.4,
      "position": [512, 0],
      "id": "333fdb0a-b94c-479d-a23d-f2fcb2e4f4e5",
      "name": "Switch"
    },
    {
      "parameters": {
        "jsCode": "// Get AI response\nconst aiOutput = $json.output;\n\n// Convert AI JSON string into object\nconst parsedResponse = JSON.parse(aiOutput);\n\n// Return clean JSON\nreturn [\n  {\n    json: parsedResponse\n  }\n];"
      },
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [256, 16],
      "id": "9c3ccfc0-6ab7-41bd-a49b-ea0432a55e37",
      "name": "output Split"
    },
    {
      "parameters": {
        "method": "POST",
        "url": "https://platform.chatsyncs.com/api/v1/whatsapp/send",
        "sendBody": true,
        "contentType": "form-urlencoded",
        "bodyParameters": {
          "parameters": [
            { "name": "apiToken", "value": "={{ $('data').item.json.api_key }}" },
            { "name": "phone_number_id", "value": "={{ $('data').item.json.phonenumber_id }}" },
            { "name": "phone_number", "value": "={{ $('data').item.json.phone_number }}" },
            { "name": "message", "value": "={{ $json.text }}" }
          ]
        },
        "options": {}
      },
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.4,
      "position": [880, -16],
      "id": "7dc4b274-26e6-40ce-a8a4-90d7fff66a7f",
      "name": "Text"
    },
    {
      "parameters": { "httpMethod": "POST", "path": "ai-agent", "options": {} },
      "type": "n8n-nodes-base.webhook",
      "typeVersion": 2.1,
      "position": [-560, 16],
      "id": "fdd3a855-56a7-4b08-9518-fbbefbc2dd8c",
      "name": "incoming Messages",
      "webhookId": "bf25e261-b177-4866-874c-85328970af5e"
    },
    {
      "parameters": {
        "assignments": {
          "assignments": [
            { "id": "c9ecbf01-7062-47d2-b8ef-2cbf37a9c1ea", "name": "api_key", "value": "22152|xxxx", "type": "string" },
            { "id": "1baf2f4a-b6da-4522-9813-983589932a07", "name": "phonenumber_id", "value": "10414101xxxx", "type": "string" },
            { "id": "3ed22ff9-e794-4944-a2e3-4d78d53f8384", "name": "phone_number", "value": "={{ $json.body.chat_id }}", "type": "string" },
            { "id": "c7492055-aed3-47df-bca6-84737568753e", "name": "name", "value": "={{ $json.body.custom_fields.Name }}", "type": "string" }
          ]
        },
        "options": {}
      },
      "type": "n8n-nodes-base.set",
      "typeVersion": 3.4,
      "position": [-320, 16],
      "id": "88ef89db-22ad-4298-a560-f3a8a49cee2e",
      "name": "data"
    },
    {
      "parameters": {
        "method": "POST",
        "url": "https://platform.chatsyncs.com/api/v1/whatsapp/send/interactive-buttons",
        "sendBody": true,
        "contentType": "form-urlencoded",
        "bodyParameters": {
          "parameters": [
            { "name": "apiToken", "value": "={{ $('data').item.json.api_key }}" },
            { "name": "phone_number_id", "value": "={{ $('data').item.json.phonenumber_id }}" },
            { "name": "phone_number", "value": "={{ $('incoming Messages').item.json.body.chat_id }}" },
            { "name": "message", "value": "={{ $('Switch').item.json.body }}" },
            { "name": "buttons", "value": "={{ $('Switch').item.json.buttons }}" },
            { "name": "button_footer_text", "value": "={{ $('Switch').item.json.footer }}" },
            { "name": "media_url", "value": "https://img.magnific.com/free-vector/people-walking-sitting-hospital-building-city-clinic-glass-exterior-flat-vector-illustration-medical-help-emergency-architecture-healthcare-concept_74855-10130.jpg" }
          ]
        },
        "options": {}
      },
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.4,
      "position": [880, 192],
      "id": "7b5f6b30-f42b-4118-a360-d5b5688d3418",
      "name": "image and interactive"
    }
  ],
  "connections": {
    "Simple Memory": { "ai_memory": [[{ "node": "AI Agent", "type": "ai_memory", "index": 0 }]] },
    "OpenAI Chat Model": { "ai_languageModel": [[{ "node": "AI Agent", "type": "ai_languageModel", "index": 0 }]] },
    "AI Agent": { "main": [[{ "node": "output Split", "type": "main", "index": 0 }]] },
    "When chat message received": { "main": [[]] },
    "Switch": {
      "main": [
        [{ "node": "Interactive", "type": "main", "index": 0 }],
        [{ "node": "Text", "type": "main", "index": 0 }],
        [{ "node": "image and interactive", "type": "main", "index": 0 }]
      ]
    },
    "output Split": { "main": [[{ "node": "Switch", "type": "main", "index": 0 }]] },
    "incoming Messages": { "main": [[{ "node": "data", "type": "main", "index": 0 }]] },
    "data": { "main": [[{ "node": "AI Agent", "type": "main", "index": 0 }]] }
  },
  "active": true,
  "settings": { "executionOrder": "v1", "binaryMode": "separate" },
  "meta": { "templateCredsSetupCompleted": true },
  "nodeGroups": [],
  "tags": []
}

This workflow only contains the n8n-side logic — it has no knowledge of your OpenAI credentials or your ChatSyncs API token. Reconnect the OpenAI Chat Model node’s credential and fill in your own apiToken and phone_number_id in the HTTP Request nodes before testing, or it will fail to reach anything.

Features Used

Feature	Purpose	Documentation
Incoming Webhook	Forwards every customer message to your AI agent’s server	Bot Settings
Send Text Message	Delivers the AI’s plain-text reply to the customer	Send Text Message
Send Interactive Buttons	Delivers the AI’s reply with tappable Reply Buttons	Send Interactive Buttons

Expected Result

A customer messages the WhatsApp number as usual. ChatSyncs forwards the message out to the company’s own AI agent, that AI agent decides what to say using its own tools and memory, and the reply lands back in the same WhatsApp conversation — as plain text or Reply Buttons — with the customer never aware the answer came from outside ChatSyncs.

Frequently asked

Why not just use ChatSyncs' built-in AI Agent instead?

ChatSyncs’ own AI Agent answers from documents trained into a Knowledge Base campaign — a great fit if you’re starting fresh. This pattern is for the opposite situation: an AI agent that already exists, already has its own tools and memory, and shouldn’t be rebuilt or retrained a second time just to reach WhatsApp.

Does ChatSyncs ever see my AI's training data or tools?

No. ChatSyncs only sees the incoming message it forwards out, and the finished reply it’s asked to send back. Everything in between — the model, the memory, the tools, the routing logic — stays on your own server.

Is this the same thing as Webhook Workflow?

No — they run in the opposite direction. Webhook Workflow is for an external event (like an order being placed) triggering ChatSyncs to send a message. The Incoming Webhook used here is the reverse: a customer message on WhatsApp triggering a POST out to your server.

Do I need my own server to build this?

Yes — the Incoming Webhook needs a real, internet-reachable endpoint to POST to, and the Send Text Message / Send Interactive Buttons calls need something to make them. n8n is one option; any server or workflow tool that can receive a POST and make one back works the same way.

What happens if my AI agent's server is down or slow?

The Incoming Webhook doesn’t wait for a response — it’s fire-and-forget, so a slow or down server just means the customer never gets a reply, not that the bot itself breaks. Add your own timeout and fallback handling (e.g. a default “we’ll get back to you” message) on your server if this matters for your use case.

Can I mix plain text and interactive buttons in the same conversation?

Yes — the Switch node in this build can send either format depending on what the AI decides fits, message by message.

Does this only work with n8n?

No — n8n is just what this build used. Any server capable of receiving the Incoming Webhook’s POST and calling ChatSyncs’ Send Text Message / Send Interactive Buttons endpoints can run this same pattern, in any language or workflow tool.

What if I want the AI to send a message the customer didn't ask for, outside the 24-hour window?

Session messages (both endpoints used here) only work within 24 hours of the customer’s last message. Sending outside that window requires an approved Message Template instead — see Send Text Message for that limit.

Bot Settings — the Incoming/Outgoing Webhook tab used in Step 1.
Send Text Message — the plain-text send endpoint used in Step 3.
Send Interactive Buttons — the buttons send endpoint used in Step 3.
AI Agents and Intent Detection — ChatSyncs’ own built-in AI Agent, for contrast.
Webhook Workflow — the opposite-direction pattern: an external event triggering ChatSyncs to send.
Finding IDs — where to get the apiToken and phone_number_id used in Step 3.

Back to Business Problem

Start over from the beginning of this use case.

​Prerequisites

​Implementation Guide

​Step 1 — Enable the Incoming Webhook

​Step 2 — Your AI agent generates a reply

​Step 3 — Send the reply back through the ChatSyncs API

​Copy & test this workflow

​Features Used

​Expected Result

​Frequently asked

​Related Documentation

Back to Business Problem

Prerequisites

Implementation Guide

Step 1 — Enable the Incoming Webhook

Step 2 — Your AI agent generates a reply

Step 3 — Send the reply back through the ChatSyncs API

Copy & test this workflow

Features Used

Expected Result

Frequently asked

Related Documentation