IDiyas

For the Inventor, By the Inventor

San Jose, CA, United States of America

Sandeep Agrawal

LinkedIn
USPTO Granted Patents = 14 


Average Co-Inventor Count = 4.5

ph-index = 4

Forward Citations = 36(Granted Patents)


Company Filing History:

goldMedal14 out of 11,294 
 
Oracle International Corporation
 patents

Years Active: 2019-2024

where 'Filed Patents' based on already Granted Patents

14 patents (USPTO):
1. 12072953 - Asymmetric allocation of SRAM and data layout for efficient matrix-matrix multiplication (2024-08-27),
2. 11989657 - Automated machine learning pipeline for timeseries datasets utilizing point-based algorithms (2024-05-21),
3. 11868854 - Using metamodeling for fast and accurate hyperparameter optimization of machine learning and deep learning models (2024-01-09),
4. 11790242 - Mini-machine learning (2023-10-17),
5. 11720822 - Gradient-based auto-tuning for machine learning and deep learning models (2023-08-08),
6. 11620568 - Using hyperparameter predictors to improve accuracy of automatic machine learning model selection (2023-04-04),
7. 11562178 - Adaptive sampling for imbalance mitigation and dataset size reduction in machine learning (2023-01-24),
8. 11544494 - Algorithm-specific neural network architectures for automatic machine learning model selection (2023-01-03),
9. 11429895 - Predicting machine learning or deep learning model training time (2022-08-30),
10. 11176487 - Gradient-based auto-tuning for machine learning and deep learning models (2021-11-16),
11. 11138291 - Assymetric allocation of SRAM and data layout for efficient matrix multiplication (2021-10-05),
12. 11120368 - Scalable and efficient distributed auto-tuning of machine learning and deep learning models (2021-09-14),
13. 10521225 - Matrix multiplication at memory bandwidth (2019-12-31),
14. 10452744 - Memory management for sparse matrix multiplication (2019-10-22),

Title: **Innovator Spotlight: Sandeep Agrawal**

Introduction

Sandeep Agrawal is a distinguished inventor based in San Jose, CA. With an impressive portfolio of 14 patents, he has significantly contributed to the fields of memory architectures and machine learning techniques. His innovative approaches, particularly in optimizing matrix multiplication and automating machine learning pipelines, reflect his commitment to pushing the boundaries of technology.

Latest Patents

Among his notable inventions, Sandeep's latest patents include:

1. **Asymmetric allocation of SRAM and data layout for efficient matrix-matrix multiplication**

This technology enhances the performance of matrix multiplication in systems utilizing scratchpad memories or associative caches. By allocating asymmetric memory space for both the left and right matrices as well as a dot product matrix, the system efficiently executes dot product operations, optimizing the use of memory resources and enhancing computational speed.

2. **Automated machine learning pipeline for timeseries datasets utilizing point-based algorithms**

This invention introduces an innovative window preprocessor that transforms training timeseries datasets for machine learning models. By generating and evaluating various preprocessor configurations, the system selects the optimal one for transforming new datasets. This capability is pivotal in enabling real-time processing of live streaming timeseries, making it a valuable asset in the field of data science.

Career Highlights

Sandeep Agrawal’s career is marked by his role at Oracle International Corporation, a major player in cloud computing and enterprise software. His work at Oracle focuses on advancing technologies that enhance data processing efficiency and performance optimization in software solutions.

Collaborations

Throughout his career, Sandeep has collaborated with talented professionals like Sam Idicula and Nipun Agarwal. These associations have fostered a creative exchange of ideas, driving innovation within their projects and leading to more robust technological advancements.

Conclusion

Sandeep Agrawal’s contributions to the fields of matrix multiplication and automated machine learning significantly reflect his innovative spirit. With 14 patents to his name and continuous collaboration with esteemed colleagues, he remains a pivotal figure in the tech industry, constantly contributing to the improvement of computational technologies and methodologies.

This text is generated by artificial intelligence and may not be accurate.
Please report any incorrect information to support@idiyas.com
Loading…