🛡️ CreadGraph AI

Enterprise Graph Neural Network (GNN) for Sybil-Resistant Talent Acquisition

Defending referral networks with deep learning and cryptographic security.

---

📌 Mission Architecture

CreadGraph AI is a production-grade Web3-inspired talent acquisition platform. It fundamentally changes how enterprises handle employee referrals by deploying Graph Neural Networks (GNN) to proactively identify, isolate, and eradicate bot-nets, sybil attacks, and referral fraud.

Designed with a high-performance FastAPI backend and a React 18 frontend, the system processes multi-nodal candidate relationships in real-time to generate heuristic Trust Scores.

🚀 Key Enterprise Features

• 🧠 Deep Learning Fraud Engine: Built using PyTorch Geometric. The system dynamically reconstructs the database into a complex NetworkX graph. Incoming users are evaluated against pretrained weights to flag abnormal clustering and high-degree bot referrals.
• 🔒 Cryptographic Security: Implements rigorous backend protection using OAuth2PasswordBearer. All passwords are mathematically hashed via Bcrypt, and sessions are governed by stateless JWT (JSON Web Tokens).
• 🌐 Real-Time Telemetry Dashboards:
  • Recruiter Matrix: Manage job mandates, track candidate funnels, and process real database entries.
  • Scout Hub: Dedicated user portals for sharing encrypted referral links and monitoring financial rewards.
  • Admin Intelligence: High-level God-View to monitor GNN accuracy, trigger manual suspensions, and observe live cluster threats.
• 💎 Premium Aesthetic Design: Leveraging Framer Motion and Tailwind CSS to deliver a zero-latency, dark-themed, glassmorphic user experience suitable for modern cybersecurity tools.

---

🛠️ Comprehensive Technology Stack

Architecture Layer	Core Technologies	Purpose
Frontend Framework	React 18, Vite	High-speed component rendering and build pipeline.
State & Fetching	Context API, Custom Hooks, Axios	Centralized authentication state and interceptor-ready API calls.
UI / UX	Tailwind CSS, Framer Motion, Lucide	Responsive design, micro-interactions, and scalable iconography.
Backend Engine	Python 3, FastAPI, Uvicorn	Asynchronous, highly concurrent API routing with automated OpenAPI specs.
AI / Machine Learning	PyTorch, PyTorch Geometric, NetworkX	Tensor calculations, synthetic graph training, and neural inference.
Data & ORM	SQLite, SQLAlchemy, Pydantic	Relational data integrity, schema validation, and fast local querying.
Security Auth	Jose, Passlib, Bcrypt	JWT token issuance and irreversible password hashing.

---

🚦 Local Environment Setup

Deploy the entire infrastructure locally in minutes.

Phase 1: Neural Engine (Backend)

# 1. Clone and navigate to the backend
git clone https://github.com/Dipanshusinghh/CreadGraphAI.git
cd CreadGraphAI/backend

# 2. Establish a secure Python virtual environment
python -m venv venv
source venv/Scripts/activate   # On Windows: .\venv\Scripts\activate

# 3. Install strictly pinned dependencies
pip install -r requirements.txt

# 4. Generate the database and populate with synthetic Graph Data
python seed_db.py

# 5. Ignite the FastAPI ASGI Server
python run_server.py

Backend Endpoint: http://127.0.0.1:8000
OpenAPI Swagger Docs: http://127.0.0.1:8000/docs

Phase 2: Client Interface (Frontend)

Open a secondary terminal window:

# 1. Navigate to the project root
cd CreadGraphAI

# 2. Install Node modules
npm install

# 3. Launch the Vite Development Server
npm run dev

Frontend Interface: http://localhost:5173

---

🔑 Initial Authentication Matrix

Upon running seed_db.py, the following encrypted credentials are automatically provisioned:

Role	Email Address	Password	Permissions
Super Admin	admin@creadgraph.ai	admin123	System oversight, User suspension, GNN monitoring.
Verified Scout	rajan@gmail.com	password	Submit referrals, view wallet, track statuses.

---

Architected for scale. Engineered for security.

Created by Deepak Singh