RahmaCarehealthcare in the family's hands
A Flutter app for managing a family's health with AI assistance: initial triage, regular follow-up, and an experience built around the person before the screen. The product is in development.
- Client
- Our own product — not a client project
- Sector
- Health and relief
- Year
- 2024 — present
- Role
- Full product: design, engineering, release
- Platform
- iOS · Android
- Status
- In development
Resultsالنتائج
What changed
The result first, then how we got there. Numbers you can trace to a source — not adjectives.
The challengeالتحدّي
The problem we came to solve
A single family manages appointments, medicines and reports for several people across paper, WhatsApp messages and the mother's memory. And in health relief specifically, polished numbers are everywhere with no traceable source.
- A health record per person with different permissions inside the same family
- An AI assistant that reads the case and warns early — without making the medical decision
- No number published on the platform without a source that can be traced
- Full accessibility: keyboard, screen reader, and a path for deaf users
Approachالمقاربة
What we decided — and why
A health file for the whole family
A record per person with their appointments, medicines and reports, with different view permissions per role inside the family. The father sees his own, the mother sees the children, the doctor sees only what was granted.
An assistant that reads the case, not just answers
An AI layer summarises the indicators and warns early, while the medical decision stays human. Every alert carries its source: which reading, which date, and which threshold it crossed.
A ledger that cannot change quietly
Every action — appointment, medicine, aid — is recorded with a serial number and a hash chained to the one before it. The beneficiary sees who viewed their file and when, and consent is explicit and revocable.
Solutionالحل
What was actually delivered
A Flutter app with one codebase for iOS and Android, over AWS with Cognito for authentication, with an AI layer that works inside the task flow rather than in a side panel. Accessibility is not an add-on: the full path is tested with a screen reader, with a dedicated path for deaf and non-speaking users.
Next step
Have a project of this size?
Tell us about the process, and we come back with a first diagnosis and a scope estimate — no commitment.
- We reply within two working days
- The code is entirely yours
- Published numbers, signed and verifiable