cv
General Information
Full Name | Himanshu Shekhar Sahoo |
Date of Birth | 19th July 1993 |
Languages | English, Hindi, Odia |
Education
- 2023
Doctor of Philosophy, Electrical Engineering
University of Minnesota (UMN), Minnesota, USA
- 2017
Masters, ELectrical Engineering
UMN, Minnesota, USA
- 2015
Bachelors, Information and Communication Technology
Dhirubhai Ambani Institute of Information and Communication Technology, Gujarat, India
Work Experience
- 2019 - 2023
Lead AI and NLP Researcher
Department of Electrical Engineering, UMN
- Developing grant proposals, writing journals, securing funding, exploring existing literature and state-of-the-art speech-to-text tools for project specific use cases, and increasing cross-collaboration with multiple teams across various institutions.
- 2021 - 2023
Deep Learning Researcher
Department of Surgery, UMN
- Using ML and NLP techniques, developing a novel clinical annotation system for high-volume, real-time extraction and analysis of acute and long-term COVID-19 symptoms from unstructured patient notes.
- Comparable to other state-of-the-art clinical annotation systems, the developed system was the quickest, utilized little computational resources, and had good symptom extraction performance.
- 2022
Health Futures Researcher
Microsoft
- Developed lesion detection pipeline to identify lesions (sprains, cartilage loss, fracture etc) present in a patient’s MRI image reconstructed using undersampled data.
- In addition to detecting rare lesions, the pipeline had similar detection performance for MRI images reconstructed using no undersampling and MRI images reconstructed using an undersampling factor of 2, 4 and 8.
- 2016 - 2021
Teaching Instructor
Department of Electrical Engineering, UMN
- Ability to work under pressure while teaching and grading assignments in more than 20 undergraduate and graduate level classes with class sizes ranging between 20-150 students.
- 2018
AI Infrastructure Data Science Intern
NVIDIA
- Developed state of the art multi-core multi-GPU ML models in CUDA C++.
- Analyzed and alleviated existing scaling inefficiencies present in these ML models.
- The developed models were a part of the NVIDIA RAPIDS library.
Grants Received
- 2021
- Cisco Tech for HealthCare Research Grant
- UMN Informatics Institute Small Seed Grant
Achievements
- 2021
- Google PhD Fellowship - University Finalist
Areas of Interest
-
GPGPU architecture optimization
-
Machine Learning
- Natural Language Processing
- Explainable ML
- Fairness in ML
Other Interests
- Hobbies: Reading, Traveling, Table Tennis, Cooking