Privacy-Preserving Analytical Pipelines Using Differential Privacy and Secure Multi-Party Computation in Federated Cloud Frameworks

Abstract

The exponential growth of data in cloud-based systems, coupled with rising concerns about data privacy, has spurred the demand for secure, privacy-preserving analytical methods. This paper proposes a federated cloud framework integrating Differential Privacy (DP) and Secure Multi-Party Computation (SMPC) into analytical pipelines. The approach enables collaborative data analytics across decentralized institutions without compromising sensitive information. By combining DP's statistical obfuscation and SMPC's cryptographic protection, the system supports privacy guarantees even in adversarial or semi-honest settings. Evaluation results demonstrate that the proposed design balances utility, privacy, and scalability—making it suitable for sectors like healthcare, finance, and smart governance.