A tutoring agency pays for a WhatsApp booking assistant
MarcoKueksHacker News · 2026-04-14
Read originalA tutoring company paid me $5K to build an AI agent that books calendar appointments over WhatsApp.
Business model
The visible HN post describes a tutoring-booking assistant, a $5,000 implementation fee and two-day delivery. It is an operator claim with a tool-promotion interest, not verified revenue.
A tutoring company buys an implementation that reduces scheduling work. WhatsApp and calendar access are dependencies, not included Muse features.
Workflow
- Collect courses, teacher availability and cancellation rules
- Convert booking conversations into proposed slots
- Write confirmed slots to the calendar and detect duplicates
Adaptation plan
The original business is not demonstrated to use Muse. The personal-agent workflow is an untested AgentFlow adaptation. Public claims are not audited results. Exact permissions, operating costs and replication performance remain unverified.
Starter prompt
Written by AgentFlow; not the original author’s prompt. Not independently deployed in Muse or Grok.
Plan a small, reviewable pilot for: A tutoring agency pays for a WhatsApp booking assistant. I will supply real business materials, budget and allowed accounts. Tasks: Collect courses, teacher availability and cancellation rules; Convert booking conversations into proposed slots; Write confirmed slots to the calendar and detect duplicates. Return drafts, missing inputs and a results table. Mark unknown facts explicitly. Do not invent customers, revenue or completed actions. Do not publish, contact people, pay or change accounts without approval. Separate proposed actions from executed ones.
Running costs and pitfalls
The original business is not demonstrated to use Muse. The personal-agent workflow is an untested AgentFlow adaptation. Public claims are not audited results. Exact permissions, operating costs and replication performance remain unverified. No independent cost benchmark is available. Check subscriptions, permissions and external actions before running a workflow.