Multi-Party Computation API

WEB-MPC is an open-source platform designed to facilitate privacy-preserving data collection and aggregation through secure multi-party computation (MPC) techniques. It enables multiple participants to submit sensitive data via web interfaces without exposing individual inputs to others, ensuring confidentiality throughout the data collection process. The platform is particularly suited for scenarios where privacy is critical, such as surveys, research studies, or any collaborative data analysis requiring strong data protection guarantees.

The core of WEB-MPC relies on the JIFF MPC library, bundled as a submodule, and requires a Node.js environment, MongoDB for data storage, and Docker for containerized deployment. It supports multiple deployments, allowing different data collection campaigns to run with distinct configurations, data templates, and HTTPS settings. Users interact with the system through a web UI to generate sessions, manage participation links, collect data, and unmask aggregated results securely.

What sets WEB-MPC apart is its web-native approach to MPC, making privacy-preserving computations accessible without specialized client software. Its modular deployment system and integration with standard web technologies simplify setup and customization. The platform includes end-to-end testing suites to verify deployments and supports secure production deployment practices, including reverse proxy setups with Nginx and Let's Encrypt. Developers can get started by cloning the repository, installing dependencies, and running the Docker-compose setup or following detailed AWS EC2 deployment instructions provided in the documentation.