Inpher XOR Platform is an AI and machine learning (ML) platform offering data science capabilities so users can combine and analyze sensitive, federated information sources without seeing or moving the input data.

Inpher XOR architecture uses a scalable secure Multiparty Computation (MPC) protocol and contains three main components:

- XOR Analyst Portal: An intuitive web graphical user interface (GUI), extensive Python library (xor-py), and REST API provide accessibility and a versatile programming environment for a wide range of users and privacy-preserving applications, from simple SQL queries to training advanced AI and ML models.

- XOR Service: Hosted by Inpher or an authorized provider, XOR Service receives and compiles the program from the Analyst Portal and orchestrates the MPC protocol among the XOR Machines (data sources). The XOR Service can’t see or store the data.

- XOR Machines (data sources): Virtual Machines installed at each data source receive instructions from the XOR Service. The XOR Machines then communicate with one another through secret shares and each send secret shares of the output data back to the Analyst Portal, which are combined to provide the final result of the program.