Svipe ID

A digital identity app - we took over the codebase from another team and shrank the app by 90%

Project: Svipe
Industry: Identity verification
Services: Mobile Development

Svipe ID is an application allowing for digital identity verification based on data read from the electronic system of a passport / ID card and verification of the user's face. On this basis a unique identifier authenticated by Authorizing Certificates is generated which allows to confirm the user's identity without the need to use a physical ID card.

Summary

The client asked us to take over the development from an existing team, identify and improve functionalities that negatively affect the performance and User Experience of the application.

Methodology

After receiving the application code and consulting with the client, we identified the main areas in need of improvement. Due to the high dynamics of changes and the need for a quick response we focused on weekly sprints. Thanks to this approach, we were able to divide the work into smaller increments and allow the client to quickly react to its effects.

Key challenges

Optimization of an underdeveloped face detection module which negatively affected the application performance
Thanks to the implementation and optimization of more modern face detection solutions, we have shortened the module operation time by several orders of magnitude.
An OCR module that slows down the performance of the application
We optimized the existing OCR module reading the text from the ID card, which resulted in a significant improvement in the speed of the module.
Reduction of the application size
The SvipeID application uses complex Machine Learning models that are responsible for face detection. Initially, the model was contained directly in the application's archive, so its size exceeded 100MB. Thanks to the optimization of the application archive and the transfer of the model to an external server, we were able to reduce the initial size of the application by 90%.

Gallery

Stack

Android
iOS

Effect

See the effects of our work on the project Svipe

Better app performance
Optimizations significantly accelerated the registration process and the general “smoothness” of the application.
Smooth operation of the camera and face detection mechanics
Currently, the app is several times more fluid and user-friendly than in the before the implemented changes.
Positive user feedback
The introduced changes were positively received by the client and resulted in positive opinions of users.

Similar projects