Building InZob: Turning WhatsApp Into an AI Engagement Layer for Healthcare and Local Business
A case study in how InZob went from a slow, manual follow-up problem to a working AI engagement and appointment platform — and the product decisions that mattered along the way.
InZob is a live, working platform for AI-powered business and healthcare engagement. The problem it solves is unglamorous but expensive: enquiries and patient leads come in across channels — website, WhatsApp, phone — and without a system, they sit unanswered until someone gets around to them. By then, a meaningful share of them are gone.
The problem, precisely
The obvious fix — "hire more people to respond faster" — doesn't scale for a small clinic or business, and a generic chatbot doesn't actually solve it either, because the hard part isn't answering messages. It's qualifying intent, routing the right conversation to the right outcome, and making sure nothing falls through after the first message.
What had to be true for this to work
- Conversations needed to happen where people already are — WhatsApp and web — not a new app nobody would install.
- AI agents had to do real qualification, not just answer FAQs — understanding what a lead or patient actually needs before a human gets involved.
- Every conversation had to become structured, trackable data — a CRM view, not a disappearing chat log.
- Follow-up had to be automatic by default, because manual follow-up was the entire original problem.
What was built
The platform combines AI-powered conversations across web and WhatsApp with agents that qualify leads and patients, a CRM layer for tracking conversation status, automated appointment scheduling, and automated follow-up sequences so nothing waits on a person to remember. On the healthcare side specifically, it extends into AI-assisted clinical documentation — patient summaries, prescription templates, and suggested doctor's notes drafted from the conversation, for a clinician to review and finalize.
The product decision that mattered most
The temptation with a project like this is to build the CRM first, because it feels like the "real" product. InZob's actual leverage point was the conversation layer — the AI agent handling first contact well enough that a human only needs to step in for the conversations that actually need one. The CRM and reporting matter, but they're downstream of that first interaction working.
That ordering — solve the highest-leverage workflow first, build the supporting structure around it — is the same pattern worth applying to almost any AI product idea that starts from a real, specific pain point rather than a feature list.