PASCAL PAILLIER

ZAMA

Learn more about building an AI-first technology startup on the Air Street Capital blog and our monthly analytical newsletter, Your Guide to AI.

The Research and Applied AI Summit (RAAIS) is a community for entrepreneurs and researchers who accelerate the science and applications of AI technology. We’ve been running for 6 years now and have hosted over fifty entrepreneurs and academics who have built billion-dollar companies and published foundational papers that drive the AI field forward. 

In the lead up to our 6th annual event that will be broadcast live online on the 26th June 2020, we’re running a series of speaker profiles highlighting what you can expect to learn on the day!

Homomorphic encryption for deep learning: a revolution in the making

At RAAIS 2019, we featured talks on privacy-preserving ML techniques such as federated learning and differential privacy from Brendan McMahan (Google AI) and Andrew Trask (OpenMined). We explored how these systems work at scale in production systems such as Google’s mobile keyboard and how they could enable future set of applications in domains where data access is sensitive, such as healthcare.

This year, we’re going deeper into the topic of privacy-preserving ML techniques. We’re excited to be hosting Dr. Pascal Paillier for a talk on fully homomorphic encryption (FHE), a form of encryption that allows computation on encrypted data (ciphertext) such that the results are the same as those running on plaintext. Several flavors of FHE have emerged in the last decade, which include parallel or sequential, scale-invariant or circuit-specific, multi-key or single-key, floating point or integer, symmetric or asymmetric homomorphic computations.

Dr. Pascal Paillier is a homomorphic encryption expert and the CTO of Zama, a Paris-based startup that builds open-source software tools enabling developers to easily deploy secure deep learning applications powered with homomorphic encryption. Zama ambitions to propel deep learning into a revolutionary new era of privacy-preserving cognition.

In its first release, the Zama toolchain focuses on homomorphic inference as a service, where unencrypted cognitive models are running in the cloud but their evaluation is performed over user-encrypted data.

You can read more about Pascal's research here, more about Zama here, and follow Zama on Twitter here. Welcome, Pascal!