Break Down Data Silos with Cryptographic Security Using Inpher on Oracle Cloud

Break Down Data Silos with Cryptographic Security Using Inpher on Oracle Cloud


Privacy concerns abound when organizations work with sensitive information. Government agencies in particular retrieve, analyze, and manage large volumes of confidential and classified data. Organizations are often hesitant to share data sets for fear of data leaks, privacy breaches, and cyberattacks. How do you gain valuable insights from your data, give customers and stakeholders the confidence that their information is kept confidential, and address applicable compliance and data protection laws? To help you solve these challenges, Oracle has added a new marketplace tool to our ecosystem that offers customers an easy-to-use, secure data collaboration platform for analytics and artificial intelligence (AI).

Introducing Inpher: An Oracle Cloud Marketplace offering

Inphernow available on Oracle Cloud Marketplace in Oracle’s public cloud regions* as well as government and defense regions in the US and UK—is a leader in privacy-enhancing computation that empowers organizations to collaborate on sensitive data seamlessly and securely across teams and borders. Inpher’s proven 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—and Inpher’s XOR Platform and related services and offerings—is a third party marketplace offering that requires Inpher license terms and conditions. Additional FedRAMP and DISA P-ATOs may be required.

For government agencies that operate siloed information systems on both modern and legacy infrastructure, XOR is groundbreaking as it enables analysts to securely compute on data located on premises or the cloud and across separate networks, countries, or classification levels—all with differing compliance requirements or privacy laws. With Inpher’s XOR Platform on Oracle Cloud, sensitive data doesn’t leave the data source with XOR’s cryptographic guarantee that no information is shared between the data sources. The analyst only sees the analytical output without revealing the data behind it.

The XOR Platform

Inpher’s XOR product 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.

For a simple explainer video on MPC and secret sharing, watch the following video.

In the following figure, the XOR Service receives a computation request from the Analyst Portal and sends the data sources a randomly generated set of numbers to mask the input data. Each XOR Machine then sends partial data to the Analyst Portal for the results to be aggregated.

Figure 1: XOR Service sends input data to XOR Machines


In total, the analyst completes five main processing steps in the XOR Analyst Portal to gain deep insights into data. The following figure provides an overview of the five steps:

1. Write function

2. Compile function

3. Distributed compute

4. Outputs broadcast

5. Result computed


Data analyst steps
Figure 2: Data analyst steps


In this accompanying video, Inpher’s Ignacio Queralt, solutions architect, breaks down the components of the XOR Platform and illustrates how easy it is to get your data project started with the XOR platform through Oracle Cloud Marketplace. With XOR, users can securely retrieve, process, and analyze data with just a few clicks.


Increased data privacy

The security of Inpher’s XOR Service and MPC Protocol allows analysts to access sensitive data while helping ensure that both Inpher and the cloud provider do not see customer data. The data doesn’t leave the source’s system, dramatically reducing the opportunity for data to be exposed. This concept is a game changer for government entities that handle sensitive and classified information.

Inpher’s one-of-kind platform encrypts data while it’s being processed with encryption in-use technology. View the following video to learn more about how Inpher’s XOR Platform protects data privacy.



The Defense Information Systems Agency (DISA) Zero Trust (ZT) Pillars identified in the ZT strategy are centered around data, with a focus on the development of a comprehensive data management strategy to protect data. Inpher’s data security capabilities enhance Oracle’s zero-trust security model and can reduce risk by protecting and preventing access to data.

Removing data silos to drive real life value

Eliminating barriers and opening up the potential to access the full scope of sensitive and disparate data sets, XOR on Oracle Cloud enables you to address mission-critical data initiatives like never before. When government organizations collect and store data in separate environments, they can use XOR to bring those data sets together for comprehensive analysis. The data doesn’t need to be duplicated in each environment when you use XOR, reducing the time and cost burden of low to high data replication.

The democratization of data that XOR on Oracle Cloud enables can significantly increase interagency data sharing. It’s far easier for agencies to share data for privacy-preserving analytics and ML projects when each organization can enable XOR access instead of building a data transfer pipeline or providing individual user access to data. Paired with the Oracle Modern Data Platform, organizations can use Inpher’s capabilities to collaborate on important mission-critical work to address social services, national security, and environmental challenges. With the XOR Platform, agencies can even create a platform for partner organizations to train their own models on agency data without seeing the data.

In-Q-Tel, a nonprofit organization focused on investing in cutting-edge technologies to enhance U.S. national security, has partnered with Inpher to support the intelligence community by helping advance their machine learning and data analytics capabilities through privacy-enhancing technologies.


Data privacy and security are crucial for government, defense, and intelligence organizations that handle sensitive and classified information. Inpher’s integration with Oracle Cloud represents a significant advancement in secure and private cloud computing. This collaboration enhances data security, making it an ideal choice for organizations with data housed across multiple clouds and on-premises environments. Inpher’s XOR Platform is available to try for free within Oracle Cloud Marketplace, where you can gain hands-on experience with Inpher’s secure data collaboration platform.

For more information, see the following resources: