WhatsApp AI replies — planned
The intended experience
Section titled “The intended experience”AI-assisted replies are intended to help a team answer common customer questions using approved business information. Human-reviewed suggestions and automatic replies are separate modes.
The planned workflow is:
- An authorized administrator supplies approved business context.
- An agent can review and edit a suggested reply before sending.
- Automatic replies are enabled separately, with explicit policy and usage controls.
- Human takeover pauses automatic replies.
- AI resumes only through an explicit authorized action.
The aim is a simple, controlled conversation—not an unrestricted bot with access to every CRM operation.
Prepare useful business context
Section titled “Prepare useful business context”Organize accurate information about your services, pricing, availability, qualification questions, support boundaries, and escalation route. Remove credentials, internal-only policies, and unnecessary customer data.
Decide which questions always need a human. Do not let an AI response create a promise your team cannot honor.
Customer follow-ups
Section titled “Customer follow-ups”Conditional customer follow-ups are also planned. They must respect replies, human takeover, consent, and current messaging rules. A team’s internal follow-up reminder is not the same as a message sent to a customer.
Different from AI Workspace
Section titled “Different from AI Workspace”AI Workspace helps employees ask read-only questions about permitted CRM data. Planned WhatsApp replies are a customer-facing messaging capability. Neither description implies that employee AI chat can change CRM records.
Contact support for release availability. No release date is promised here.