JustifyAI

Intelligent Fact-Checking and Content Enhancement Platform

---

---

Transform AI responses into verified, trustworthy content with intelligent fact-checking and visual enhancement

Overview

JustifyAI is a comprehensive platform that combines LLM auditing with intelligent content enhancement. It systematically verifies AI-generated responses against real-world information and transforms them into visually appealing, trustworthy content.

Problem Statement

• AI models generate inaccurate or misleading information
• Users lack systematic verification of AI-generated content
• Raw AI responses lack visual appeal and proper structure
• No reliable method to validate AI claims against factual data

Solution

JustifyAI provides an automated fact-checking layer that:

• Analyzes AI responses for verifiable claims
• Verifies information using web search and knowledge bases
• Generates detailed audit reports with accuracy assessments
• Enhances content with visual formatting and images
• Rewrites responses to correct inaccuracies

Key Features

Intelligent Fact-Checking

• Multi-Agent Architecture: Specialized critic and reviser agents
• Web Search Integration: Real-time verification using Google Search
• Claim Extraction: Automatic identification of verifiable statements
• Accuracy Scoring: Detailed accuracy assessments and reporting

Content Enhancement

• Visual Formatting: Structured, engaging content presentation
• Image Generation: Relevant visuals to accompany responses
• Professional UI: Clean, modern interface with smooth animations
• Responsive Design: Seamless experience across all devices

Developer Integration

• RESTful API: Easy integration with existing applications
• Cloud-Native: Built on Google Cloud Platform for scalability
• Real-time Processing: Fast response times with streaming capabilities
• Comprehensive Logging: Detailed monitoring and debugging support

Technology Stack

Frontend

• Next.js 15.2.4 - React framework with App Router
• TypeScript - Type-safe development
• Tailwind CSS - Utility-first styling
• Framer Motion - Smooth animations
• Radix UI - Accessible component library
• Firebase - Authentication and real-time features

Backend

• FastAPI - High-performance Python web framework
• Google Vertex AI - Advanced AI and ML capabilities
• Google Cloud Run - Serverless container platform
• Pydantic - Data validation and serialization
• Uvicorn - ASGI server for production

AI & ML

• Google ADK - Agent Development Kit
• Google Gemini - Large Language Model
• Multi-Agent System - Specialized fact-checking agents
• Web Search Tools - Real-time information verification

Quick Start

Prerequisites

• Node.js 18+ and npm/pnpm
• Python 3.10+
• Google Cloud Platform account
• Google Cloud CLI installed

Installation

1. Clone the Repository

git clone https://github.com/your-username/genexhange1.git
cd genexhange1

2. Backend Setup

cd backend/llm-auditor
poetry install

# Set up environment variables
export GOOGLE_GENAI_USE_VERTEXAI=true
export GOOGLE_CLOUD_PROJECT=your-project-id
export GOOGLE_CLOUD_LOCATION=us-central1

# Authenticate with Google Cloud
gcloud auth application-default login

3. Frontend Setup

cd frontend
npm install
npm run dev

4. Run the Application

# Backend (Terminal 1)
cd backend/llm-auditor
poetry run python main.py

# Frontend (Terminal 2)
cd frontend
npm run dev

Visit http://localhost:3000 to access the application.

How It Works

graph TD
    A["User Input"] --> B["LLM Auditor API"]
    B --> C["Critic Agent"]
    C --> D["Web Search Verification"]
    D --> E["Accuracy Assessment"]
    E --> F["Reviser Agent"]
    F --> G["Content Enhancement"]
    G --> H["JustifyAI Processing"]
    H --> I["Visual Enhancement"]
    I --> J["Final Response"]
    
    style A fill:#e1f5fe
    style J fill:#c8e6c9
    style C fill:#fff3e0
    style F fill:#fff3e0

Loading

Process Flow

1. Input Processing: User submits query or statement for verification
2. Claim Extraction: Critic agent identifies verifiable claims
3. Web Verification: System searches for supporting evidence
4. Accuracy Assessment: Claims evaluated and scored for accuracy
5. Content Revision: Reviser agent corrects inaccuracies
6. Visual Enhancement: JustifyAI adds formatting and structure
7. Final Output: User receives verified, enhanced content

Use Cases

Educational Content

• Verify historical facts and scientific claims
• Enhance learning materials with visual aids
• Create trustworthy study resources

News & Media

• Fact-check articles and reports
• Generate verified news summaries
• Create engaging visual content

Business Applications

• Verify product information and claims
• Enhance marketing content with credibility
• Generate trustworthy customer communications

Research & Development

• Validate research findings
• Enhance technical documentation
• Create reliable reference materials

Example Output

Before Processing

"Earth is further away from the Sun than Mars."

After JustifyAI Processing

FACT-CHECKED RESPONSE

INACCURATE CLAIM DETECTED
Original: "Earth is further away from the Sun than Mars."

CORRECTED VERSION
"Mars is further away from the Sun than Earth."

VERIFICATION DETAILS
• Earth's average distance: ~150 million km
• Mars' average distance: ~228 million km
• Accuracy Score: 0% (Original) → 100% (Corrected)

SOURCES VERIFIED
• NASA Solar System Data
• Astronomical Almanac
• Multiple scientific databases

API Documentation

Query Endpoint

POST /query
Content-Type: application/json

{
  "message": "Your text to verify",
  "user_id": "optional_user_id",
  "generate_image": true
}

Response Format

{
  "response": "Verified and enhanced content",
  "session_id": "unique_session_id",
  "justifyai_response": {
    "formatted_content": "Enhanced visual content",
    "images": ["generated_image_urls"],
    "accuracy_score": 95,
    "verification_details": {...}
  }
}

Deployment

Google Cloud Run

# Deploy backend
cd backend/llm-auditor
poetry install --with deployment
python3 deployment/deploy.py --create

# Deploy frontend
cd frontend
npm run build
# Deploy to your preferred platform

Docker Support

# Build and run with Docker
docker build -t justifyai-backend ./backend/llm-auditor
docker run -p 8080:8080 justifyai-backend

Contributing

We welcome contributions through:

1. Bug Reports: Report issues and bugs
2. Feature Requests: Suggest new features and improvements
3. Code Contributions: Submit pull requests for enhancements
4. Documentation: Help improve documentation and examples

Development Setup

# Fork the repository
git clone https://github.com/your-username/genexhange1.git
cd genexhange1

# Install development dependencies
cd backend/llm-auditor
poetry install --with dev

cd frontend
npm install

# Run tests
cd backend/llm-auditor
python3 -m pytest tests

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Acknowledgments

• Google Cloud Platform for infrastructure support
• Google ADK for the agent development framework
• Next.js Team for the React framework
• FastAPI for the Python web framework
• All Contributors who help improve JustifyAI

---

Star this repository if you find it helpful

Made with dedication by the JustifyAI Team

Contact