Jinesh Dineshbhai Patel

Title: Jinesh Dineshbhai Patel and Multimodal Asset Quality Management.

Introduction

Jinesh Dineshbhai Patel is an inventor associated with San Jose, California, and Roku, Inc. His patent work includes a system for assessing metadata quality to support content selection.

Latest Patents

Patel is associated with 1 patent: “Asset Quality Management Using Multimodality.” The invention describes systems, methods, and software that evaluate metadata elements for use in a content-selection graphical user interface. A large language model generates embeddings for training metadata, and similarity scores between embeddings help identify metadata labeled as “good” or “bad.” A deep neural network is then trained through supervised learning to generate quality labels or scores for metadata.

Career Highlights

Patel’s disclosed patent work focuses on applying machine learning to metadata evaluation. The approach combines large language models and deep neural networks to help assess the quality of information used in content selection.

Collaborations

Patel’s listed coworker is Fei Xiao. The available information identifies both as associated with Roku, Inc., but does not specify their roles or the nature of their collaboration.

Conclusion

Jinesh Dineshbhai Patel’s patent describes a machine-learning approach to evaluating metadata quality, connecting language-model embeddings with neural-network-based scoring for content selection.