Intuit is investing in adopting the new techniques of fully homomorphic encryption, which are just now moving from the theoretical and purely academic sphere into the realm of practitioners, a transition that will take many years. FHE means arithmetic operations can be calculated on the encrypted data, which opens up new possibilities for privacy-safe machine learning.

Tzvika Barenholz and Induprakas Keri detail some of the progress Intuit has made in solving the many technical problems that still plague FHE, including performance, support for different types of models and AI techniques, and the related system design.

What you'll learn

• Learn how Intuit is solving some of FHE's problems to bring it to more users

This session is from the 2019 O'Reilly Artificial Intelligence Conference in San Jose, CA.