VeriAI

AI-powered verification platform for detecting misinformation and validating claims using LLMs and trusted sources.

Overview

VeriAI helps users verify the credibility of information by combining large language models with trusted external sources. Given a claim or article, VeriAI analyzes the content, retrieves supporting evidence, and generates an explainable confidence score.

Features

• 🔎 Fact-check natural language claims
• 🤖 AI-powered reasoning
• 📚 Evidence retrieval
• 📊 Confidence scoring
• ⚡ Fast API responses
• 🖥️ Modern web interface

Tech Stack

• Frontend: Flutter
• Backend: Flask
• ML models: HuggingFace hallucination models
• LLM Models: OpenAI, Meta, Gemini
• Deployment: Vercel

Architecture

User
   │
Frontend
   │
Backend API
   │
───────────────
│ LLM Service │
│ Retrieval   │
│ Database    │
───────────────

Installation

git clone https://github.com/OmarCoderBro/VeriAI.git
cd VeriAI

Install dependencies

npm install
# or
pip install -r requirements.txt

Run locally

npm run dev
# or
python app.py

Usage

1. Enter a claim.
2. VeriAI searches for supporting evidence.
3. The AI analyzes credibility.
4. Review the confidence score and explanation.

Screenshots

Add screenshots or a demo GIF here.

Roadmap

• Browser extension
• PDF verification
• Multi-language support
• Citation export
• User accounts

Contributing

Contributions are welcome! Feel free to open an issue or submit a pull request.

License

MIT License