We’re excited to share highlights from Dr. Jianshu Weng’s recent session at the Open Data Science Conference (ODSC), one of the world’s premier gatherings for the global data science community. At this renowned AI training conference, which brings together innovative minds from across the field, Dr. Jianshu, Head of Federated Learning at AI Singapore, showcased their groundbreaking work on Synergos – a federated learning platform built on top of OpenMined’s PySyft library.
In his talk, Dr. Jianshu highlighted a critical challenge in modern AI development: the need to balance data access with privacy concerns. While machine learning models typically improve with more data, organizations face increasing regulatory pressures from frameworks like GDPR and PDPA when sharing sensitive information.
Enter Synergos, AI Singapore’s innovative solution that leverages PySyft’s core capabilities to enable privacy-preserving machine learning across multiple organizations. By choosing PySyft as its foundation, AI Singapore demonstrates the growing trust in OpenMined’s open-source tools for handling sensitive data at a national level.
The ODSC presentation showcased how Synergos extends PySyft’s functionality to make federated learning more accessible to end-users, abstracting away complex technical details while maintaining robust privacy guarantees. This approach allows organizations to collaborate on AI models without exposing their underlying data – a crucial feature for both regulatory compliance and protecting commercial interests.
We’re particularly excited that AI Singapore chose to build upon OpenMined’s technology stack. Their decision validates our commitment to creating secure, privacy-preserving AI tools that can operate at scale. We eagerly welcome other government organizations interested in similar implementations to reach out and explore how our open-source tools can support their privacy-preserving ML initiatives. Please reach out to @LaceyStrahm on our community Slack.
