MISSION BRIEF / AI
SA7BY — Egyptian Virtual Assistant
An Android graduation project that turns Egyptian-dialect voice commands into device actions.
Explore architecture ↓Graduation project. This overview follows Mohamed’s supplied description; the repository is private and the demonstration opens in Google Drive.
MY ROLE
University graduation project development.
The problem
Interpret Egyptian-dialect voice commands and map them to useful actions on an Android device.
Architecture
- Voice commandSpeech recognition
- Speech recognitionIntent processing
- Intent processingServer response
- Server responseAndroid action
- Android actionWorkflow output
Engineering decisions
- Use speech-to-text recognition to accept spoken Egyptian-dialect commands.
- Classify recognized text into intents with a machine-learning workflow.
- Separate server-side interpretation from Android device actions through a returned response.
Challenges
- Understanding Egyptian-dialect speech and mapping different commands to intents.
- Coordinating speech recognition, server-side processing and the response on the device.
The solution
An Android application using speech recognition and natural language processing. After a trigger phrase, the command is recognized and processed through a server-side intent workflow. A response returns to the device to select the corresponding action.
The result
The graduation project demonstrates voice-driven actions such as setting an alarm, opening applications, making calls and searching. A demonstration video is linked below; no accuracy figures or production-scale results are claimed.