2024-07-12 –, North Hall
Fully Homomorphic Encryption (FHE) is a privacy-enhancing technology that enables performing computations over encrypted data. FHE has recently seen a lot of progress, and commercial applications of FHE are now available. One of the main application domains for FHE is privacy-preserving machine learning. We introduce a Python interface for OpenFHE, a popular open-source FHE C++ software library that supports all common FHE schemes. OpenFHE is a NumFocus-sponsored open-source project that has been authored by a community of well-known FHE cryptographers and software engineers.The talk provides a high-level introduction to FHE and its applications, and then provides an overview of the Python API. Several examples are presented to both illustrate FHE concepts and show the practicality of the technology.
More information about the OpenFHE project:
* Main website: https://www.openfhe.org/ ;
* OpenFHE discourse forum: https://openfhe.discourse.group/ ;
* Main OpenFHE repository: https://github.com/openfheorg/openfhe-development ;
* OpenFHE organization: https://github.com/openfheorg ;
* Main OpenFHE design paper: https://eprint.iacr.org/2022/915 ;
Advanced
Contributor to the OpenFHE project and Machine Learning Lead
Sukanya is an experienced Research Engineer and Tech Lead, specializing in driving innovation in the Data and AI field. She has led teams in developing Federated Learning use cases across various domains. Her work involves extensive research and development in Federated Learning and Edge AI, creating prototypes, solutions and architecting and developing data systems for diverse industrial domains.
She is also an organizing committee member of PyCon Ireland and led organizer of PyData Ireland.