Sonal Shukla

Experience

• 

  Grasim Industries Limited | Pulp & Fibre

  May 2017 - Jul 2017

  Internship Trainee
• 

  National Remote Sensing Centre, ISRO

  Jun 2018 - Nov 2018

  Project Intern

  I had the incredible opportunity to collaborate with esteemed researchers at ISRO, where I embarked on a transformative journey. I extend my heartfelt gratitude to our mentor, Dr. Vinod Bothale, for providing me with this enriching internship experience characterized by unwavering professionalism.During this tenure, our focus was on DevOps, with a special emphasis on Selenium, a ubiquitous tool in the realm of automated testing across diverse industries. Our overarching goal was to delve into the various facets of testing methodologies and employ a range of cutting-edge tools to automate processes on a demonstration website. This initiative ultimately formed an integral part of a classified project, showcasing the real-world impact of our efforts. Show less

• 

  University at Buffalo

  Mar 2020 - now

  As a Research Assistant at Dr. Scott Doyle's lab within the Department of Computational Pathology and Anatomical Sciences at UB, I am passionately engaged in pioneering research at the intersection of artificial intelligence and healthcare.My primary focus is on leveraging Federated Learning techniques to advance the field of Computational Pathology, specifically in the realm of Whole Slide Images (WSIs). My mission is to develop robust Deep Learning models for segmentation and classification tasks, particularly in the context of Papillary Thyroid Carcinoma and Oral Cavity Cancer WSIs. What makes this endeavor unique is our ability to achieve these objectives without the need to share sensitive medical data.The inspiration driving my work stems from the potential impact it could have on healthcare. Imagine AI models capable of predicting patient outcomes, an innovative development that could profoundly assist medical practitioners and patients alike in making informed decisions regarding prognosis and treatment strategies.Moreover, I am proud to share that both my PI and I recently secured a grant from CTSI, NIH, amounting to $50K. This grant will fuel our pilot study for the upcoming year, allowing us to delve deeper into our research and further our mission.Join me on this journey at the cutting edge of AI and healthcare, where innovation is not only transforming our understanding of diseases but also enhancing the quality of care for patients worldwide. Show less Project title: Enhancing Workplace Safety: A Deep Learning Approach to Real-time Hand Gesture Recognition and Traffic Light ControlDuring my independent research study, conducted under the guidance of Dr. Chunming Qiao, I had the privilege of delving into the exciting field of Hand Gesture Recognition through the application of advanced Deep Learning algorithms. The primary objective of this research was to leverage real-time camera input for the detection of human gestures in a dynamic working environment, with a particular focus on enhancing workplace safety by triggering a responsive traffic light system.My role encompassed various critical tasks, beginning with the meticulous collection and curation of a specialized dataset tailored to the intricacies of hand gestures. Subsequently, I embarked on the development and implementation of Deep Learning models. These models were designed to not only recognize and interpret hand gestures accurately but also to interact seamlessly with the surrounding infrastructure to initiate appropriate responses, such as controlling traffic lights.This enriching experience not only honed my technical skills in the realm of computer vision and Deep Learning but also fostered a deeper understanding of the real-world applications of artificial intelligence. Show less Project title: Extracting Explainable Text Information from CuesAs an Independent Study student, I conducted research to assist individuals with impulsive decision-making tendencies, such as those facing challenges like obesity. My primary objective was to modify ingrained behaviors by enhancing subjects' ability to delay gratification. To achieve this, I focused on measuring and improving their Delay Discounting (DD) capabilities.Key Accomplishments:*Collaborated with Dr. David Doermann, an esteemed mentor, to develop and execute this research initiative.*Applied Machine Learning (ML) techniques to identify cues that could positively impact subjects' decision-making processes.*Employed advanced NLP algorithms to identify words and linguistic patterns that could enhance Delay Discounting and promote better decision-making.*Contributed to the development of a robust framework for improving the ability to delay gratification, thereby facilitating behavior modification goals.*Conducted extensive data analysis and experimentation to assess the effectiveness of various cues and words in influencing subjects' decision-making behaviors.This experience not only expanded my expertise in ML and NLP but also allowed me to make a meaningful contribution towards empowering individuals to make more informed and deliberate choices, ultimately leading to improved outcomes in behavior modification. Show less This project was a "Parameter Study of COVID19" using pattern recognition under Dr. Dong. The outbreak of coronavirus proved to be a massive threat to global public healthwith adverse impacts on the socio-economic status of affected countries. Predictive modelinghelps essentially to identify the crucial preventive measures to undertake the flattening ofpandemic spread. Traditional SIR (Susceptible-Infectious-Recovered) model has been a basis forunderstanding the paradigm of epidemics. Still, it cannot undertake many critical metrics anddenies certain vital aspects like the evaluation of the incubation period, lockdown and socialdistancing measures, asymptotic infections, and several stages of an infected population. Theaddition of these key metrics also plays a crucial role in optimizing the observations based on thenotion of a traditional model. The massive rate of transmission of the novel coronavirus deniesunderlying assumptions of the classical kinetic systems, henceforth the report discusses amodified epidemic model to predict and analyze the changing trend of the epidemic situation,then analyze the parameters involved in the infection dynamics model. Acknowledging thatCOVID-19 situation has been varying from time to time, several assumptions false out over aperiod with new reports getting published; it is challenging to model realistic situations with noassumptions; however, the dataset that's being acted upon makes a significant difference inevaluating the epidemic. The report discusses some of the key metrics that can help in uniquelyidentifying the epidemic situation without relying on many assumptions. Show less

  • Graduate Research Assistant

    Jul 2021 - now

  • Graduate Student

    Jan 2020 - now

  • Independent Study

    Aug 2020 - Dec 2020

  • Independent Study

    Jul 2020 - Dec 2020

  • Project Student

    Mar 2020 - Sept 2020
• 

  Iron Mountain

  May 2024 - now

  Digital Pathology Intern in AI