WhatsApp automation
I help connect customer conversations with the systems and people that can resolve them. WhatsApp is one place that work can begin.
WhatsApp automation can combine an enquiry, an approved business-system action, and a handoff to a person. The right design depends on the information available, the action the business allows, and how exceptions are handled.
Discuss on WhatsAppConnect the next step
A useful assistant needs to know when to answer a question, retrieve information, or pass the conversation to your team. A message that sounds helpful is not the same as completing the underlying task.
For a proposed engagement, we can map common questions, approved information sources, existing tools, and the people responsible for escalations. System access, messaging requirements, and implementation scope must be agreed before connecting a live channel.
Relevant project work
My Shifa work brings hospital enquiries into WhatsApp. The reviewed implementation includes appointment flows, lab-report delivery, AI voice-note replies, and reception handoffs. The code connects WhatsApp, Nuacare hospital workflows, and Chatwoot.
This describes reviewed implementation scope. Production rollout, individual delivery responsibilities, and outcomes are still being confirmed. The public WhatsApp mockup and welcome audio are illustrative and contain no patient conversation.
I also led Progbiz Connect for Progbiz, a product focused on AI agents across WhatsApp, voice, and web conversations.
Explore the Shifa preview and scopeExplore Progbiz ConnectPlan the boundaries
The assistant should have a specific job and a clear limit. Sensitive actions and unexpected requests need an explicit path for review.
- Identify what the assistant can answer from approved information.
- Map which actions need access to an existing business system.
- Define how a conversation reaches the right person.
- Test missing information, failed integrations, and requests outside scope.
Before we talk
Can it use existing systems?
That depends on the system’s API, permissions, and the workflow. The Shifa implementation illustrates connections between WhatsApp, Nuacare, and Chatwoot; it does not establish compatibility with every product.
Can it include voice notes?
AI voice-note replies are part of the reviewed Shifa implementation. The portfolio includes a synthetic welcome preview. Languages, delivery requirements, and quality for your own use case need to be confirmed.
Will people still be involved?
Human handoffs are part of the approach. We should define the issues an assistant can handle, the cases a person should review, and the team responsible for follow-up.