Skip to main content

Project Structure

The EdgeAI project follows a well-organized structure that separates concerns and makes the codebase maintainable.

Directory Overview​

EdgeAI/
├── app/                          # Android application
│   ├── src/main/
│   │   ├── java/                 # Kotlin/Java source code
│   │   ├── cpp/                  # Native C++ code
│   │   └── assets/               # App assets
│   └── build.gradle.kts          # Build configuration
├── docs/                         # EdgeAI Documentation Platform
│   ├── docs/                     # Documentation content
│   ├── src/                      # EdgeAI source files
│   ├── static/                   # Static assets
│   └── docusaurus.config.ts     # EdgeAI configuration
├── backup_qnn_libs/              # QNN library backups
└── build/                        # Build outputs

Key Components​

Android App (app/)​

  • Kotlin/Java: Main application logic and UI
  • C++: Native inference engine with ExecuTorch + QNN
  • Assets: Model files and resources

Documentation (docs/)​

  • EdgeAI Documentation Platform: Modern documentation framework
  • Organized content: Technical docs, setup guides, releases
  • Responsive design: Mobile-friendly documentation

Native Libraries (backup_qnn_libs/)​

  • ARM64-v8a: Optimized for modern Android devices
  • QNN libraries: Qualcomm Neural Processing SDK
  • Version-specific: v79 context binaries for SoC Model-69

Architecture Benefits​

  1. Separation of Concerns: Clear boundaries between app, docs, and libraries
  2. Maintainability: Easy to locate and modify specific components
  3. Scalability: Structure supports future expansion
  4. Documentation: Self-documenting through organized docs structure

Development Workflow​

  1. App Development: Work in app/ directory
  2. Documentation: Update files in docs/docs/ directory
  3. Native Code: Modify C++ files in app/src/main/cpp/
  4. Build & Test: Use Gradle build system
  5. Deploy: Automatic documentation deployment via GitHub Actions