Rouben Sagatelov

Title: Rouben Sagatelov and Federated Learning

Introduction

Rouben Sagatelov is an inventor associated with Highland Heights, Ohio, and United Services Automobile Association (USAA). His patent work addresses federated learning, a method that can help improve machine-learning models while keeping data collected by participating devices local.

Latest Patents

Sagatelov is associated with 1 patent: “Endpoint Device Peer Ecosystem for Federated Learning.” The invention describes a network of lower-capability endpoint devices, such as devices in a home or office, that collect user, usage, or environmental data and communicate with a local central device.

The central device can receive a machine-learning model from a remote computing system. It may collect and aggregate data from the endpoint devices to retrain the model, or distribute the model so the endpoint devices can retrain it individually. The central device can then aggregate the resulting models and return an updated model to the remote system without exposing the peer ecosystem’s collected data.

Career Highlights

Sagatelov’s listed patent connects his work with USAA and focuses on coordinating machine-learning tasks across endpoint devices and a local central device. The approach is designed to support model improvement while limiting the transfer of underlying data.

Collaborations

Mark Anthony Lopez and Dustin Bowen Bitter are identified as Sagatelov’s coworkers. The available information does not specify their roles in the patent or whether they collaborated on this invention.

Conclusion

Rouben Sagatelov’s patent describes a federated-learning ecosystem that enables local devices to contribute to model training while keeping their collected data from being shared with a remote computing system.