CASE STUDY
EFS Attend
About the project
EFS Attend is a mobile attendance management application built for field-based and on-site teams.
The app enables team leads to manage their teams by adding users and registering facial data directly within the application.
Attendance is marked using face recognition by capturing a single image, eliminating the need for manual check-ins.
Face recognition is integrated using on-device machine learning to ensure fast and reliable performance.
The application follows an offline-first approach, allowing attendance to be recorded even without an internet connection.
All offline data is securely stored in a local database for up to 45 days and automatically synced when connectivity is restored.
Location-based geofencing ensures attendance can only be marked within approved work areas.
To prevent misuse, the app blocks attendance marking if the device time is not properly synchronized.
The solution is developed in Flutter and supports both Android and iOS platforms.
Key achievements
- Implemented face recognition–based attendance using TensorFlow Lite and ML Kit with approximately 99% accuracy.
- Designed and implemented an offline-first architecture with local data retention for up to 45 days.
- Built seamless background synchronization to upload stored attendance data once internet access becomes available.
- Integrated geofencing to restrict attendance marking to authorized physical locations.
- Added device time validation to prevent attendance fraud caused by manual time changes.
- Developed complete team management workflows allowing team leads to add users and register facial profiles.