Deeksha Dangwal

Title: Deeksha Dangwal: Innovator in Neural Network Technologies

Introduction

Deeksha Dangwal is a prominent inventor based in Santa Barbara, CA. She has made significant contributions to the field of neural networks, particularly in the development of software that enhances the functionality of hardware neural network processors. Her innovative approach has led to advancements that bridge the gap between quantized firmware and software implementations.

Latest Patents

Deeksha holds a patent titled "Deriving a concordant software neural network layer from a quantized firmware neural network layer." This patent outlines systems and methods for deriving a software neural network layer that is compatible with a quantized firmware layer. The method involves processing data using a neural network processor (NNP) to generate quantized data, followed by matrix-vector multiply operations to produce results that can be replicated by different processors.

Career Highlights

Deeksha is currently employed at Microsoft Technology Licensing, LLC, where she continues to push the boundaries of technology in neural networks. Her work has been instrumental in developing systems that improve the efficiency and effectiveness of neural network applications.

Collaborations

Throughout her career, Deeksha has collaborated with notable colleagues, including Jeremy Halden Fowers and Daniel Lo. These partnerships have fostered an environment of innovation and creativity, leading to groundbreaking advancements in their field.

Conclusion

Deeksha Dangwal's contributions to neural network technology exemplify her dedication to innovation and excellence. Her patent and work at Microsoft Technology Licensing, LLC highlight her role as a leading inventor in the tech industry.