Aug 2022 – Aug 2024 · 2 yrs

• Advisors: Dr. Amit Sharma

Jan 2022 – Jul 2022 · 6 mos

• Advisors : Dr. Sunayana Sitaram, Dr. Monojit Choudhury

Mar 2021 – Apr 2022 · 1 yr 1 mo · Paris, Île-de-France, France

Feb 2021 – May 2021 · 3 mos · Cambridge, Massachusetts, United States

Sep 2020 – Feb 2022 · 1 yr 5 mos · New Delhi, India

• Working in the field of Artificial Intelligence in Healthcare.
• Worked on Optimal Vaccine allocation using Deep Reinforcement Learning Methodologies such as ACKTR, Deep Q Network Model and Contextual Bandits using features like Death Rate, Susceptible, Population, Recovery Rate, etc to make an optimal allocation of resources on the basis of these ever-changing features. The main aim is to ensure that the overall count of people susceptible to COVID-19 gets decreased thus mitigating the after affect and preventing the virus from spreading.
• Got featured as part of the Spotlight Presentations at MIT’s Conference “Vaccines for All”
• Worked on a research project Mining Trends of COVID-19 Vaccine Beliefs on Twitter with Lexical Embeddings, focusing on identifying and tracking the temporal trend of public perception and influencing factors behind it on vaccine uptake using lexical categories on Twitter posts.

Aug 2020 – May 2021 · 9 mos · Jundiaí, São Paulo, Brazil

• Umgrauemeio (formerly known as Sintecsys) is a Brazilian Agritech company recognised as one of the top "100 Climate innovators" by Epoca Negocios magazine, working on the problem of Automatic Wildfire Detection using Computer Vision and Deep Learning techniques to provide fast detection solutions and stop the fire from producing serious damages. I worked on creating algorithms for early wildfire detection and reducing CO2 emissions originating from it contributing to severe global warming. My work at Sintecsys includes but is not limited to :
• Worked with satellite imagery and on-ground live image analysis to form algorithms for optimising the reaction time of detecting fire outbreak and obtaining information for better preventive measures.
• Performing extensive Data Analysis and pre-processing of the images obtained through cameras deployed on the plantations under surveillance.
• Optimising and implementing various deep learning models to obtain the state of the art results for Classification and Segmentation on the images captured by the high-resolution cameras deployed by Sintecsys on the Plantations and Forest areas.
• Working on various features with an aim to reduce the time taken to detect wildfire as soon as it breaks out and perform accurate localisation by forming a bounding box.

May 2020 – Jul 2020 · 2 mos · Jaipur, Rajasthan, India

• Piltover Technology is an upcoming organisation which aims at developing affordable high-end products catering to the needs of the disabled people belonging to the economically weaker section of the society. My work at Piltover Technologies included:
• Implementing a host of Machine Learning models from scratch for gesture classification using Electromyography signals generated in the limbs to be embedded in fully automated prosthetics.
• Performing all the required tasks of preprocessing the raw EMG signals, implementing muscle activity detection and extracting the time domain based features for classification.
• Reconstructed and deep dived into several relevant research papers to create an efficient classification model using time domain specific features.

Apr 2020 – Dec 2020 · 8 mos · Mumbai, Maharashtra, India

• Working on an interactive chatbot to register harassment reports and provide the correct resources to the victims to deal with the situation. The chatbot will aim at getting the victims in touch with the NGOs and correct authorities to deal with the problem and will be hosted on popular social media platforms like Facebook and WhatsApp so that it is easily available to the people in need.
• Developed a fully detailed Anti Street Harassment poll to understand the surge in Domestic Abuse and Online Harassment during the period of lockdown. Performed in-depth analysis on the collected data using Python and Google Data Studio to help the Red Dot Foundation team optimize their outreach
• Using Data Analytics techniques to gain hindsight of the increased Domestic Abuse that women face in India during the lockdown phase, using applications of Twitter Scraping, Natural Language Processing and Inferential Statistics.
• Performed analysis on reported crimes against women in Pune using Data Science and Analysis techniques. Google Data Studio and Python were used for creating Statistical Visualisation of the analysis obtained.
• Part of the 'IAMCOMINGOUT' project by Red dot Foundation and The Jimme Foundation where we aim to build a public platform for Queer(LGBTQ+) people to seek assistance from. The public repository will help the LGBTQ+ community to expand and get in touch with organisations all over the country which can help them with any mental, economical, social issue they are struggling from
• Deep dived into the coursework taught in our country to understand and formulate an inclusive curriculum regarding sex education to make the students more aware and knowledgable. Developed insightful dashboards using GDS in order to understand the possible educational reforms required to inculcate Sex Education.
• Developed First Respondent module to increase effective inclusivity and awareness for the LGBTQ+ community regarding topics like Empathy building and Allyship.