> For the complete documentation index, see [llms.txt](https://doc.aissist.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://doc.aissist.io/tutorial/deploy/wati.md).

# WATI

Deploy Aissist on WhatsApp through WATI (beta) - connect with an API token, create the webhook, choose the numbers and the AI operator.

*Last updated: November 6, 2026*

Deploy Aissist on WhatsApp through WATI, so it answers chats on the WhatsApp numbers you choose and hands over to your team. Read [Deploy](/tutorial/deploy.md) first for the steps every gateway shares.

{% hint style="info" %}
The WATI gateway is in beta.
{% endhint %}

### Get your WATI credentials

1. Sign in to your WATI workspace and go to **API Docs** (under **Connectors → API**).
2. Copy the **Access Token**.
3. Copy the **API Endpoint** (your tenant URL, such as `https://live-mt-server.wati.io/123456`).

{% hint style="warning" %}
Treat the access token like a password. Never share it in chat or tickets.
{% endhint %}

### Connect WATI

1. Go to **Deploy → Gateways**, click **Add Gateway** and choose **WATI**.
2. On **Connect**, choose a WATI connector, or create one: paste the API endpoint into **Tenant API URL** and the access token into **API token**. Leave **API V3 URL** at its default.

### Create the webhook in WATI

The **Connect** step shows the gateway's webhook URL. It looks like `https://gateway.aissist.io/gateway/wati/endpoint?gatewayId=...`.

1. In WATI, go to **Connectors → Webhooks** and click **Add Webhook**.
2. Paste the webhook URL from Aissist.
3. Select the event **Message received (BSUID)**.
4. Set the status to **Enabled** and save.

Back in Aissist, click **Next**. Aissist checks that the webhook exists, is enabled and has the right event; you cannot continue until it does.

Optionally, set a **Handover Operator Email**: the WATI operator a chat is assigned to when Aissist escalates. If empty, the chat goes to the WATI bot.

### Set up the agent

* **Mode**: **Auto-pilot** sends WhatsApp replies automatically (WATI has no draft mode). **Only process if assigned to AI** makes Aissist skip chats that are not assigned to the AI operator.
* **AI Operator**: the **AI Operator Email** decides which chats are Aissist's, and an optional **Display Name**.
* **Channels**: the WhatsApp numbers Aissist manages, loaded from your WATI account. By default, all.

### Additional steps in WATI

* **Assign chats to the AI operator.** Create a WATI automation (or use WATI's AI agent connection) so new chats are assigned to the AI operator email.
* **Mind the 24-hour window.** WhatsApp allows free-form replies only inside an open customer session, so Aissist answers chats while the customer's session is open.

### Actions

The same WATI connector can carry these actions on the chat. See [Add Actions](/tutorial/add-actions.md).

* **Close** the chat. WATI marks it solved.
* **Reopen** the chat.
* **Add tags** that already exist in WATI.

These run about a minute after the reply.

### Do it with Ask Engine

> Is our WATI gateway deployed, and which WhatsApp numbers does it manage?

Ask Engine answers from the gateway's settings, including whether it is deployed.

### Best practice

* Start with one WhatsApp number in **Channels**.
* Use a dedicated operator email for Aissist, and assign new chats to it with a WATI automation.
* Set a **Handover Operator Email** so escalated chats reach a person, not the bot.

### Related

* [Deploy](/tutorial/deploy.md)
* [Escalation and Handover](/capabilities/escalation-and-handover.md)