Interface for neural network to recognize any watches by photo

Design Process

Even though we did not know all the possibilities 
of the neural network, we wanted to develop smooth user experience that would work like magic. 
Therefore, our design process was a little bit different.

  1. Define
  2. Ideate
  3. Prototype
  4. Test


We ran the ideation workshop with a project team. 
We started with the question:

Who has experience with services based on neural network?
How positive is it?

The goal was to find out the process that is used in recognition by photo services and is familiar to users.

We realized that people will not give a second chance 
if no results were found.

We decided that in MVP we will check the quality of the neural network, thus we will be able to highlight the blockers which user might experience.

In the next version, we will focus on the ways of monetization and improving recognition quality.


We have described the possible recognition results

We worked out the user scenarios and general structure 
of the mobile application based on network capabilities.

User Flows


During the prototype stage we only used Native iOS controls for fast development and to be flexible for upcoming changes.

This solution helped us to release the MVP in 2 weeks.

Test & Release

Although we didn’t have time for usability test, we did some quick testing with internal team for basic usability issues and 
in app communication. We were able to understand that the results screen was really important as it provided clear in-app communication.

App Store release screens

Results and the client Success

Next Steps

The next steps would be 
to make more rigorous user testing with customers, 
so we can get more insights.

Additionally, we planned 
to build web-app to drive more traffic.


How we did it

No items found.

How we did it


No items found.

More Success Stories