Blockchain technology offers a secure and decentralized platform for financial transactions, particularly in cryptocurrency and token ecosystems. This paper introduces a real-time fraud detection system that utilizes a risk scoring model, providing a streamlined and efficient alternative to traditional machine learning methods. The system evaluates key parameters such as transaction size, frequency, velocity, geographical location, and Know Your Customer (KYC) status. Using Solidity-based smart contracts, risk scores are computed directly on-chain, enabling immediate flagging of high-risk transactions without the need for centralized oversight. This approach eliminates the latency and computational overhead associated with off-chain processing while maintaining blockchain’s transparency and data integrity. The proposed system demonstrates superior fraud detection accuracy, scalability, and adaptability, making it highly suitable for decentralized finance and other high-transaction environments. By integrating real-time monitoring and automated risk evaluation, it ensures a proactive response to emerging threats. Future directions include refining risk parameters, exploring external data integration, and enhancing system performance for broader financial applications.