Integrations · WhatsApp Business

WhatsApp AI agent integration

Run Nexora Chat as a WhatsApp Business agent through the official Cloud API. The agent handles inbound and outbound conversations, sends template messages for transactional notifications, accepts image and document uploads from customers, and supports catalogue + payment-link flows. Sessions persist across days; customers can pause and resume without losing context. Pre-approved WhatsApp Business Service Provider (BSP) connections are supported alongside direct Cloud API deployments.

How it works

Three steps to ship WhatsApp Business in your pilot

STEP 1

Register your WhatsApp number

Connect your WhatsApp Business Cloud API number to Nexora, or point an existing BSP relationship at our agent endpoint. Phone-number display name and business profile sync from your Meta Business Manager.

STEP 2

Author template messages

Pre-approved template messages handle outbound notifications, appointment reminders, and transactional updates. Templates submit for Meta approval through the same admin console; the agent picks the right one based on intent and context.

STEP 3

Persist conversations across days

Inside the 24-hour customer service window the agent replies freely; outside that window template messages re-open the thread. The conversation history persists so customers resume without restating their issue or order.

What teams use it for

Common WhatsApp Business use cases

  • Customer support on WhatsApp Business with image upload for damaged-parcel claims and document upload for ID verification.

  • Order tracking and delivery callbacks via WhatsApp, with template-message reminders and pause-resume sessions.

  • WhatsApp commerce flows with catalogue messages, add-to-cart, and payment links — all without leaving the conversation.

Code preview

WhatsApp Business — send template message

HTTP
POST /v17.0/{phone-number-id}/messages
{
  "messaging_product": "whatsapp",
  "to": "<customer-msisdn>",
  "type": "template",
  "template": {
    "name": "appointment_reminder",
    "language": {"code": "en"},
    "components": [
      {"type": "body", "parameters": [
        {"type": "text", "text": "<provider-name>"},
        {"type": "text", "text": "<appointment-time>"}
      ]}
    ]
  }
}

Stylised example — not a runnable script. Endpoints, payload shapes, and authentication are documented in the integration runbook shipped during the pilot.

Frequently asked questions

Common questions

Does Nexora support the official WhatsApp Business API?

Yes. We deploy against the WhatsApp Business Cloud API (Meta's preferred path) and against pre-approved BSPs (Business Solution Providers) when customers already have a BSP relationship. Both paths give you the full template-message, media, catalogue, and payment-link surface.

Can customers upload photos to the WhatsApp agent?

Yes. Photos, PDFs, and voice notes are supported. Common flows: damaged-parcel claims (photo upload), document verification (ID upload), and prescription requests (script photo). The agent reads the media, attaches it to the customer record, and routes it where it needs to go in your back-end.

How do template messages work for outbound notifications?

Outbound notifications outside the 24-hour customer service window must use pre-approved templates per Meta's policy. Nexora ships a starter pack (appointment reminder, delivery status, payment due, order confirmation) and supports custom templates that route through your Meta Business Manager approval flow.

Is WhatsApp commerce (catalogue + payments) supported?

Yes. The agent can send catalogue messages with product items, accept add-to-cart actions in-conversation, and dispatch payment links through your payment provider. The customer never leaves WhatsApp to complete a purchase — useful for replenishment, low-friction reordering, and small-ticket retail.

Ready to wire WhatsApp Business into your pilot?

Pilot in 14 days. Live in 30.

A 30-minute scoping call. We bring the WhatsApp Business integration patterns. You bring the system credentials and the first use case.