Kaiyi Liu

Title: Kaiyi Liu and Verifiable MLOps for Autonomous Environments.

Introduction

Kaiyi Liu is an inventor based in Toronto, Canada. His patent work addresses machine-learning operations, including ways to validate models developed through collaboration.

Latest Patents

Liu is associated with 1 patent: “Verifiable MLOps to Train ML Models on Autonomous Environments.” The invention describes providing a first machine-learning model to a collaboration platform and receiving a second model based on it. The system tests the second model against criteria associated with the first model to determine whether it is valid, then publishes the second model to a repository if it passes validation.

Career Highlights

Liu works at Red Hat, Inc. His patent reflects an approach to model development that combines collaboration with verification before a model is published.

Collaborations

Liu’s coworker Marcelo Amaral is a woman. The patent description also highlights collaboration through a platform where a model can be developed from an existing model and evaluated before publication.

Conclusion

Kaiyi Liu’s patent focuses on making collaborative machine-learning development more verifiable by testing derived models before they are added to a repository.