🛡️ NeuralGuard — Private Offline Mobile Security Framework

"NeuralGuard is a state-of-the-art, local-first mobile threat detection app for Android. Running offline AI models via NDK C++ runtimes, it intercepts security vulnerabilities on-device with zero network footprint."

---

⚡ The Recruiter Takeaway (Why This Matters)

1. On-Device LLM Inference: Executes GGUF threat models locally on-device using optimized NDK/C++ runtimes and Room local databases.
2. Dynamic UI/UX: Designed using Jetpack Compose, showcasing real-time reasoning overlays and security threat alerts.
3. High-Privilege Interception: Implements secure Android Accessibility Services (Neural Shield UI parser) and Notification Listeners (Signal Watch).

---

🏗️ Architecture Design

graph TD
    UI[Jetpack Compose UI] --> VM[NeuralGuard ViewModels]
    VM --> Domain[Domain Layer - UseCases]
    Domain --> Data[Data Layer - Repositories]
    Data --> Room[(Room DB - Scans & Logs)]
    Data --> Engine[Neural Engine - C++/NDK]
    Engine --> Models[Offline GGUF Models]

Loading

---

🛠️ Quick Launch

1. Requirements

• Install Android Studio (Ladybug or newer).
• JDK 17+ and Android SDK 34 (Target SDK).

2. Setup Command

git clone https://github.com/kalyan-1845/NeuralGuard.git

Open the workspace in Android Studio and run the Gradle sync to download NDK dependencies.