How it works

The distributed data science network.

Bitfount is a secure network for sending algorithms to data instead of the other way around. Data remains behind its custodian's firewall, with enterprise-grade access controls over which analyses can be executed.

Bitfount - Federated Analytics and Machine Learning Platform

How does it work?

The underlying idea

Bitfount is built on the paradigm of distributed data science. Algorithms are sent to data, where it lives, rather than the other way around.

  • Datasets are safely deployed within Processors of Data (Pods).
  • Data providers choose disclosure controls defining what analyses data scientists can run.
  • Data scientists’ models or queries are sent over the network for processing and checked against the data providers' rules.
  • Data scientists get what they need. No more, no less.

Want to understand the communication architecture more? Read our blog article on why we chose a messaging architecture instead of a more standard client-server approach.

Bitfount - Federated Analytics and Machine Learning Platform
Bitfount - Federated Analytics and Machine Learning Platform

Advanced disclosure controls

Before data analyses are sent back to the data scientist, Pods check the analysis against the approved disclosure controls. Bitfount's controls enable:

  • The latest privacy enhancing technologies (PETs), including differential privacy and secure aggregation, to minimise the risks of disclosure.
  • Usage restriction beyond which data is used.
  • Easily granted and revoked privileges via a simple UI.
  • Audit logs of all access controls.

Want to understand more? Read our resources on PETs → and usage-based access controls →

How we deploy

We've worked hard to make it easy to get up and running without compromising on functionality and security.

  • Get going in minutes with our python library or docker image.
  • Maintain zero-trust security with OAuth and SAML authentication.
  • Deploy within your own infrastructure or on the cloud.

Read more about our zero-trust deployment approach on our blog →

Bitfount - Federated Analytics and Machine Learning Platform