● Automated the extraction and normalization of earnings call transcripts utilizing Python scripts and NL TK.● Engineered a sentiment analysis pipeline using a pre-trained FinBERT model and 50,000+ data points from varied sources.● Designed and deployed predictive models to analyze stock price fluctuation, utilizing heat maps and time series analyses forthe financial data of Fortune 500 companies for the past 5 years to predict stock prices with an F1 score of 92%.● Currently annotating the largest online dataset to predict hawkish or dovish stance of the Central Bank using FOMC data.

Working on revolutionising BNPL payments under the Fusen Fellowship 2024 over the summer.

Splitit provides installment payment services to merchants who want to enable buy now pay later services to their customers.1) Recently began working as a SWE Intern at Splitit USA’s Atlanta Office working in the Product Division.2) Reporting directly to the Head of Product 3) Using Python based libraries, Jupyter Notebooks, and Agile Methodology to work on team-based projects.

Accepted to the 2023 cohort of the Google Computer Science Research Mentorship Program (CSRMP) under which I will be working with a Google Research mentor for the next 4 months on "Defining a Research Problem" and will receive invaluable guidance on my ongoing research at Georgia Tech's SocWeb Lab on the THRIVE and REDUCE machine learning projects.

Working with Dr. Munmun De Choudhury in the SocWeb (Social Dynamics and Wellbeing) Lab (School of Interactive Computing) led by Dr. Choudhury and working on the THRIVE and REDUCE projects.THRIVE, is a collaboration with clinical researchers and practitioners at Northwell Health in New York. It aims to examine how machine learning and computational techniques may be applied to patient-contributed social media and other online data to eventually inform clinical decision-making and positively influence treatment in the wild.REDUCE, aims to develop improved public health surveillance algorithms for real-time estimation of rates of suicide nationally. Unfortunately, suicide rates have risen nearly 30% nationally over the past 15 years. Planning for suicide prevention activities in this emerging epidemic is limited by a lack of real-time trend data on the suicide burden. Consequently, public health organizations like the CDC are investigating the accuracy and applicability of novel approaches that rely on using machine learning models to combine data from multiple time series data sources to more accurately estimate the real-time burden of suicide and self-directed violence nationally.

Teaching Assistant for CS 2050: Intro Discrete Math CS1) Responsible for leading a study group of 50 people and holding weekly teaching sessions.2) Responsible for holding office hours every week to clear doubts and offer personal assistance to students.3) Responsible for grading homework, exams, and other assignments.4) Learnt core leadership and team building skills.

ASDL’s GT Smart Campus team of research faculty and students collaborates with GT Facilities engineers to evaluate data from campus buildings, central plants, and other infrastructure, including the ultra-sustainable Kendeda Living Building. Students joining this team will help out with data processing, analytics, and machine learning to better understand and inform the performance of campus energy (and water) systems.

Data Science Research Intern.

Published research paper and Python based research internship.