Validating Healthcare AI in a Zero Trust Environment: 1-hour Workshop

BeeKeeperAI, Inc.

A guided walkthrough on the fundamentals of validating machine learning model inferences in a confidential computing environment.

In this workshop, you will gain an understanding of:

  • Working within a zero-trust and confidential computing environment
  • The methods and tools required to prepare an algorithm and data for validation

Zero trust enables secure collaborations between data stewards and algorithm developers utilizing leading-edge technology to protect both data and intellectual property.

Within a zero-trust environment:

  • The data never leaves the data steward environment
  • The data is never seen by anyone other than the data steward
  • The algorithm is never seen by anyone other than the developer

The BeeKeeperAI platform operates in the Azure Confidential Computing cloud and is accessed through two secure portals customized for the algorithm developer and the data steward. Within Azure, BeeKeeperAI enables an encrypted algorithm to be sent to the data steward’s HIPAA-compliant environment where the encrypted data and algorithm meet in an Intel SGX secure enclave for validation.

To operate in a zero-trust environment, the data steward and the algorithm developer must leverage fundamental tools and methods to optimize their collaboration including:

  • Algorithm factoring
  • Data curated against a data specification
  • End-to-end encryption of the algorithm and data
  • Establishing containers in Azure