Premium private beta Human approval for every AI-assisted send

Supervised WhatsApp AI for racket clubs

Answer with the player, booking and club context already in view — then let a person approve the exact message before it leaves WhatsApp.

No. It prepares a draft, but a human must explicitly approve every outbound WhatsApp message before it is sent.

One player record across every workflow

Conversations Live operations
Conversation detail panel showing the day’s complete WhatsApp message turns

AI-assisted draft

CRM context and club procedures shape the draft before your team reviews it.

Review before send

01 · WhatsApp

Start with the inbound message and contact

A direct message is matched to the CRM by phone. For a new, non-spam contact, the workflow can create a basic player record from the WhatsApp display name and phone, so staff have a record to continue from rather than an isolated chat.

Daily WhatsApp conversation list linked to CRM players

02 · CRM

Review CRM facts and prior conversations

The draft can use the linked player’s contact details, preferred language, public-facing profile notes, Academy leads, current or upcoming Courses, recent Playtomic bookings and day-by-day conversation history. Staff can open Dashboard → Players → the player to verify that context before replying.

Player profile Conversations tab showing linked WhatsApp history
Workflow

How the supervised reply workflow works

The sequence below covers the operational path from inbound message to draft, approval, send, retained history and follow-up. A draft is only a proposal: the operator can revise or reject it, and only explicit approval releases the outbound message.

  1. 01

    Receive and classify the inbound message

    The workflow resolves the direct-message phone. A known phone uses its CRM player; a new non-spam contact can receive a basic CRM record, while sales or spam should not create customer follow-up.

  2. 02

    Confirm the linked player

    Check name, phone and relevant CRM facts. For an inbound customer turn, context is locked to that contact rather than searching unrelated players.

  3. 03

    Read recent turns, summaries and Academy context

    Review the recent conversation context available to the assistant, including today’s unsummarized turns, plus the player’s relevant Academy leads.

  4. 04

    Review and revise the proposed reply

    Treat the assistant text as a draft. Check facts, tone, customer language, prices and promised action; request changes or reject it when needed.

  5. 05

    Explicitly approve the exact outbound text

    Only the human approval decision releases the pending WhatsApp message. Rejection sends nothing, and no autonomous reply is permitted.

  6. 06

    Keep the exchange in conversation history

    Inbound and approved outbound turns append to the day record. Staff can later read the open day’s turns or its saved daily summary from Conversations or the player profile.

  7. 07

    Create the supervised operational follow-up

    The operator can create or update a task, create a meet, or record an Academy lead. A linked lead task is optional and created deliberately by staff. Customer inbound turns cannot perform these operator-only actions themselves.

Keep communication and actions under operator control

Inbound customer turns cannot use operator-only meet, task or Academy lead workflows. The player’s CRM language stays in the generation context, outbound text remains pending until a person approves it, and staff decide whether the request should become operational work.

Turn the conversation into owned work

After the reply, the operator can create or update a task, create a meet, or record an Academy lead. Tasks linked to a lead are optional and created deliberately by staff. These remain operator-supervised decisions, not autonomous actions triggered by the customer message.

Task slide-over with assignee, due date, linked record and checklist
Academy lead slide-over with course and schedule preferences

Frequently asked questions

Supervised WhatsApp AI for racket clubs

Answer with the player, booking and club context already in view — then let a person approve the exact message before it leaves WhatsApp.

Request AI beta access