Santa Clara, CA, United States of America

Rohit Prakash Prabhavalkar

This inventor holds 39 USPTO granted patents and 7 published patent applications and 14 EPO patents, primarily in Automatic Speech Recognition. Top assignees: Google Inc., Other. Active years: 2016-2026.

IDiyas Innovation Intelligence. (2026). Inventor Profile: Rohit Prakash Prabhavalkar. Retrieved from https://idiyas.com/inventor/rohit-prakash-prabhavalkar

Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile

USPTO Granted Patents = 39 

% Patents Active = 97.4

 

Average Co-Inventor Count = 6.1

ph-index = 5

Forward Citations = 82(Granted Patents)


Location History:

  • Mountain View, CA (US) (2023)
  • Santa Clara, CA (US) (2016 - 2024)
  • Palo Alto, CA (US) (2024)

Company Filing History:


Years Active: 2016-2026

Loading Chart...
Loading Chart...
Areas of Expertise:
Automatic Speech Recognition
Contextual Biasing
Sequence-to-Sequence Models
Key Phrase Spotting
Two-Pass Recognition
Neural Network Models
Long-Form Training
Phoneme-Based Contextualization
Word Error Rate Training
Decentralized Misrecognition
Streaming ASR
Modular Encoder Features
39 patents (USPTO):Explore Patents

Title: **Rohit Prakash Prabhavalkar: Innovator in Automated Speech Recognition**

Introduction

Rohit Prakash Prabhavalkar is an esteemed inventor based in Santa Clara, CA, renowned for his significant contributions to the field of automated speech recognition (ASR). With a remarkable portfolio of 25 patents, he has advanced the technology that underpins how machines understand and process human speech.

Latest Patents

Among his latest innovations are two notable patents that enhance the functionality of ASR systems. The first patent, titled "Cascaded Encoders for Simplified Streaming and Non-Streaming ASR," describes an ASR model comprising a first encoder, a second encoder, and a decoder. This model efficiently processes sequences of acoustic frames to generate a higher order feature representation, streamlining the speech recognition process.

The second patent, "Speech Recognition with Sequence-to-Sequence Models," outlines a method for performing speech recognition utilizing sequence-to-sequence models. This method includes the reception of audio data for an utterance and the processing of features indicative of the utterance's acoustic characteristics. By leveraging an encoder and an attender to generate context vectors, the system effectively produces speech recognition scores and transcriptions, ultimately enhancing user interaction with technology.

Career Highlights

Rohit's career highlights include significant tenures at Google Inc., where he contributed to groundbreaking advancements in speech technology. His achievements have positioned him as a leading expert in the creation of innovative ASR solutions.

Collaborations

Throughout his career, Rohit has collaborated with distinguished professionals in the field, including Tara N. Sainath and Chung-Cheng Chiu. These partnerships have enriched his work and fostered a culture of innovation within the teams he has been a part of.

Conclusion

In summary, Rohit Prakash Prabhavalkar's contributions to the field of automated speech recognition through his patents and collaborations underscore his role as a pivotal inventor. His work not only enhances technological capabilities but also sets the stage for future developments in how machines interpret and communicate human speech.

Profile summary based on public USPTO records.
Data Sources: USPTO Patent Grant XML, Patent Center, EPO & CIPO • Normalized by IDiyas Innovation Graph. Methodology & provenance architecturePlease report any incorrect information to [email protected]
Loading…