Andrei Andrusenko

Title: Andrei Andrusenko: Inventor in Speech-to-Text and Conversational AI

Introduction

Andrei Andrusenko is an inventor based in Yerevan, Armenia, and works at NVIDIA Corporation. His patent work concerns speech processing and language models for conversational AI systems.

Latest Patents

Andrusenko is associated with 1 patent: “Speech-to-text processing assisted with language models for conversational AI systems and applications.” The patent describes techniques for training and deploying speech-augmented language models to capture and process speech inputs efficiently. A speech model processes audio in a first language to create part of an input to a language model, while text context associated with the audio forms another part. The language model then returns an output that includes a speech-to-text conversion of the audio.

Career Highlights

Andrusenko’s disclosed invention addresses how speech inputs and contextual text can be combined in language-model workflows. His work at NVIDIA Corporation connects his professional background with the development of conversational AI technologies.

Collaborations

The provided information identifies Zhehuai Chen and He Huang as Andrusenko’s coworkers. Their shared professional context is NVIDIA Corporation.

Conclusion

Andrei Andrusenko’s patent highlights an approach to integrating speech models, contextual text, and language models for speech-to-text processing in conversational AI.