🚀 SeiAgentPay

📌 Problem Statement

Traditional payment systems lack transparency, efficiency, and security, especially when it comes to stablecoin payments. Businesses and users face challenges such as:
⚡ High transaction fees
🔗 Lack of interoperability
🐢 Slow processing times

💡 Solution

SeiAgentPay is a decentralized stablecoin payment system that allows users to make payments seamlessly using DAI, USDT, and USDC. It integrates an AI-powered recommendation system to suggest the best stablecoin based on transaction history and user behavior, ensuring cost-effective and efficient transactions.

✨ Features

✅ 🔍 Stablecoin Recommendation – AI-powered system to suggest the best stablecoin for payments.
✅ 💰 Multi-Stablecoin Support – Supports DAI, USDT, and USDC transactions.
✅ 🔐 Secure Transactions – Trustless and transparent payments using smart contracts.
✅ 🏪 Merchant Payments – Enables direct stablecoin transfers to merchants.
✅ ⚡ Fast & Low-Cost – Built on Sei Network for high-speed and low-fee transactions.
✅ 🔧 Admin Control – Allows updating stablecoin addresses and managing settings securely.

🏗 Stablecoin Recommendation Model

SeiAgentPay integrates a machine learning-based stablecoin recommendation model that suggests the best stablecoin for a given transaction based on:

• User Transaction History (average volume & frequency)
• Merchant Category
• Transaction Purpose

Model Training & Deployment

• The RandomForestClassifier is used to train a model with transaction data.
• The model is exposed via a FastAPI backend that provides an API for recommendations.
• Encoders ensure categorical data (merchant category & transaction purpose) is processed correctly.
• The recommendation engine runs on Google Colab for model training and can be deployed locally or on a server.

🚀 Getting Started

1️⃣ Clone the Repository

git clone https://github.com/jitendragangwar123/SeiAgentPay
cd SeiAgentPay  

2️⃣ Install Backend Dependencies

cd stablecoin-recommender
pip install -r requirements.txt  

3️⃣ Train the Stablecoin Recommendation Model (Optional)

If you want to retrain the model, run:

python train_model.py  

This generates updated .pkl model files and user transaction history.

4️⃣ Run the Backend API

python main.py  

5️⃣ Install Frontend Dependencies

cd front-end  
npm install  

6️⃣ Run the Frontend

npm run dev  

📡 API Endpoints

The backend provides the following endpoints:

Endpoint	Method	Description
/	GET	Check API status
/users	GET	Fetch user transaction data
/recommend	POST	Get the best stablecoin recommendation

🤝 Contributing

We welcome contributions from the community! 🌍
Feel free to open issues, submit pull requests, or suggest new features.

📜 License

This project is licensed under the MIT License. 📝

Contact Us

Please reach out to us at rajeebk.malik@gmail.com or jitendragangwar2498@gmail.com