Omid Fatemieh

Title: Omid Fatemieh and Text Embedding Generation

Introduction

Omid Fatemieh is an inventor based in Bellevue, Washington, and is associated with Google Inc. His work includes a patent on generating text embeddings and applying them in machine-learning workflows.

Latest Patents

Fatemieh is named on 1 patent, titled “Text Embedding Generation and Applications.” The invention describes a method for responding to a user’s query by selecting an appropriate text-embedding model from multiple models, generating embeddings for data stored in a warehouse, and saving those embeddings there. It also describes using the embeddings to train a machine-learning model and notifying the user when training is complete.

Career Highlights

The patent reflects work on connecting data-warehouse operations with machine-learning model training. Its approach lets users request embedding generation and model training through queries, while the system selects a model and carries out the requested steps.

Collaborations

Fatemieh’s listed coworker is Xi Cheng. Both are associated with the patent “Text Embedding Generation and Applications.”

Conclusion

Omid Fatemieh’s patent focuses on making text-embedding generation and machine-learning training part of an integrated data-warehouse workflow.