Why this existed is easy to state and hard to fix in real life.
In my region, many health providers still maintain records that live on paper or in isolated EMR systems that do not talk to each other, so when someone moves hospitals or arrives in an emergency at an unfamiliar health facility, the story of their care is incomplete. MedSync was imagined as a calmer, connected layer: one trusted view of the patient across all health facilities, with room for the messy human parts like appointments, labs, conversations, and (carefully bounded) help from an AI assistant.
The build sits alongside academic work on what a centralised EHR would need to mean in Nigeria. This not only includes software, but adoption, policy, and day-to-day operations. The prototype is the attempt to ground that research in something you can click through.
What it actually delivers
You can see the shape of the product in this walkthrough; different roles see only what they need: hospital setup and oversight, clinical depth for doctors, lab workflows, and a patient-facing slice for longitudinal records and messaging. Chat is there because care is coordinated, not only documented. The assistant is there as a support act for doctors, as one of the major challenges for them is data entry. The assistant context is anchored on unified patient records rather than generic advice.
Next steps
A system like this only matters if it can stay online, audited, and aligned with regulation and clinical safety, especially anywhere AI touches decisions. Long term, the boring wins would be FHIR standards for interoperability, clear consent, and more role-centric workflows. These would be the next lift if or when I move it beyond prototype.
Credit to Emmanuel Tanimowo for documentation and engineering support on the journey.