Veritas AI - Content Integrity & Plagiarism Detection System

Veritas AI is a production-grade, full-stack application designed to detect AI-generated content and plagiarism with high precision. It combines traditional linguistic analysis (TF-IDF), semantic search (FAISS embeddings), and deep learning classifiers (RoBERTa) to provide a comprehensive content integrity score.

🚀 Key Features

• Multi-Layered Detection:
  • Linguistic Analysis: Feature-based detection using Perplexity (GPT-2) and Burstiness.
  • AI Classification: Transformer-based detection using a fine-tuned RoBERTa model.
  • Plagiarism Check: TF-IDF for exact matches and Semantic Search (SentenceTransformers + FAISS) for paraphrasing.
• Explainable AI (XAI):
  • Provides natural language explanations for every verdict.
  • Breaks down the "Why" behind the score (e.g., "Low perplexity indicates machine-generated variance").
• Modern UI/UX:
  • Sleek Next.js frontend with Tailwind CSS and Framer Motion.
  • Real-time analysis with visual feedback.
  • Responsive plagiarism and AI confidence visualizations.
• Unified Deployment:
  • Single-command deployment where FastAPI serves both the REST API and the static React frontend.

🛠️ Tech Stack

Backend

• Framework: FastAPI (Python)
• ML/AI: PyTorch, Transformers (Hugging Face), Scikit-Learn, Spacy
• Vector Search: FAISS (Facebook AI Similarity Search)
• Architecture: Modular Service-Repository pattern with Singleton model loaders.

Frontend

• Framework: Next.js 14 (App Router)
• Styling: Tailwind CSS
• State/Net: React Hooks, Axios
• Visualization: Recharts, Framer Motion

📦 Installation & Running

Prerequisites

• Python 3.9+
• Node.js 18+

1. Setup Backend

# Clone the repository
git clone https://github.com/Sangeeth62880/AI-Plagiarism-Checker.git
cd AI-Plagiarism-Checker

# Install Python dependencies
pip install fastapi uvicorn spacy scikit-learn sentence-transformers faiss-cpu torch transformers

# Download Spacy model
python -m spacy download en_core_web_sm

2. Build Frontend

cd frontend
npm install
npm run build
cd ..

3. Run Unified Application

python -m backend.app

The application will be available at: http://localhost:8000

🧩 Architecture

graph TD
    Client[Next.js Frontend] -->|REST API| API[FastAPI Backend]
    API -->|Route| Services
    subgraph Detection Services
        TF["TF-IDF Plagiarism"]
        SEM["Semantic FAISS"]
        AI["AI Detector (RoBERTa)"]
    end
    Services --> TF
    Services --> SEM
    Services --> AI
    TF & SEM & AI --> Fusion[Score Fusion Layer]
    Fusion --> Explain[Explainability Engine]
    Explain --> API

Loading

🛡️ License

This project is licensed under the MIT License.