Kinclara

The problem

When a parent comes home from hospital, the coordination lands on the family — usually in a WhatsApp group. Medication changes, follow-up appointments, who is collecting what, who called the GP. Group chats are built for conversation, not for state. The important detail scrolls away, and one person quietly ends up carrying all of it.

The sharpest version of this is the first fortnight after a discharge, when the instructions are densest and the family is least organised.

What I built

Kinclara gives the family one shared, private place for the practical work of care: medications, appointments, tasks, notes and shared visibility. Setup is designed to take about twenty minutes, starting from the discharge letter the family already has in their hand.

Two deliberate constraints shaped the product. It coordinates the family, not the care — no medical decisions, no diagnosis, no monitoring. And the person at the centre decides who sees what, rather than everything being visible to everyone by default.

Alongside the self-serve flow, families can book a twenty-minute concierge setup with me and I do it with them. That has been the fastest way to learn where the product actually breaks.

How it is built

React and TypeScript on the front end, Firebase Authentication and Firestore for identity and data, with services on Vercel and Railway. The whole thing was migrated off a low-code platform onto this independent stack — a full rebuild rather than a port, done to remove a dependency that would have capped what the product could become.

Product analytics run in an EU PostHog project, so behavioural data stays in the region.

Where it stands

Kinclara is live and in free private beta. Core features are free to use and will stay that way. The current work is acquisition — specifically, reaching families at the discharge moment rather than trying to interest them in caregiving software in the abstract.