The model was one part of the experience
An AI capability needs an application around it. People need to know how to capture an input, what is happening during analysis and where to find the result. The product also needs a useful response when a network connection drops or processing cannot complete.
Our role was to connect those moments into a working product, in collaboration with the client’s team.
From product decisions to the application
We delivered product and interface design, application architecture and mobile engineering for iOS and Android. The application connected guided camera capture with backend processing, then brought the resulting output back into the user’s workflow.
This required coordination across the interface, device behavior and service contracts. Capture, upload and analysis were separate operations, but the user needed to understand them as one task.
Working through the details that affect use
Camera behavior differs across devices. Our engineering work addressed capture processing and orientation differences between iOS and Android, with focused regression coverage for how captured regions were mapped.
We also worked on upload recovery and the waiting experience. Transient upload failures could be retried, while processing states, time limits and error feedback helped the application respond when a scan did not complete normally. Result-screen transitions were refined alongside those changes.
These details matter because an app has to handle more than the successful demonstration. A person using it needs a clear response when the device, network or processing service behaves differently than expected.
Delivery continued after the first build
The work included testing, release preparation and support as the product evolved. We collaborated with the client’s team to carry product decisions through implementation and investigate issues across the application workflow.
Our contribution described here concerns the application and its integration with AI processing. Model training, clinical performance and regulatory clearance are separate from this account of product delivery.
What the engagement brought together
The result was a mobile and tablet application spanning capture, processing and results, supported by a team familiar with both the product experience and the implementation behind it.
For companies developing AI-enabled medical software, that is a substantial part of the work. A model output becomes useful through the surrounding product, its integrations and the ability to support the next release.
Explore our SaMD product development services or AI development and automation.
This anonymous account describes ViewCo Digital’s contribution. References and further detail are available on request.