ABOUT ME

I am a third year undergraduate data science student at Drexel University's College of Computing and Informatics. I have experience with a breath of machine learning and development technologies and methodologies. I have worked at a Business Intelligence developer at Thomas Jefferson University Hospital for my first coop and as a Data Scientist a for my second coop. I am an aspiring T-shaped data scientist with focus on natural language processing, machine learning and their application to healthcare and other domains. I am a teaching assistant for DSCI 521 (Data Analysis and Interpretation), a graduate data science course and have been a TA for DSCI 511 (Data Acquisition and Pre-Processing) another graduate data science course and CS 570 (Programming Foundations) graduate introduction to object oriented programming. I have worked under Dr. Erjia Yan at the Scholarly Communications Lab as part of my STAR Scholars project in which we analyzed gender bias in peer reviewed publications using NLP . I love to cook and read about food in my free time. I am reading KT Achaya's A Historical Dictionary of Indian Food and Dan Jurafsky's The Language of Food: A Linguist Reads the Menu right now. I also love playing squash and watching Indian movies, my favorite is Gangs of Wasseypur by Anurag Kashyap . I also enjoy hindi poetry, साये में धूप by दुष्यंत कुमार is my favorite.
Download CV

PROJECTS


Text Processing Working Group

Working group to learn an apply NLP concepts to real world problems.

Summary

Text Processing Working Group is a student led initiative to introduce more people to Natural Language Processing and its applications. Weekly meetings focus on understanding theoretical underpinnings of concepts followed by application into real world and deployment. Open to students from all majors encouraging extensive interdisciplinary collaboration.

Go to the project

Content Based Course Recommendation System

Recommend similar courses based on course descriptions. Used TF-Idf to represent words as vectors and then used cosine similarity to get top 9 similar courses.

Summary

Recommend similar courses based on course descriptions. Used TF-Idf to represent words as vectors and then used cosine similarity to get top 9 similar courses. Link to github repository

Go to the project

Qlik Engine API Connector

Python host to interact with Qlik Sense enterprise. Uses web sockets to interface with the REST API. Created as a business intelligence developer at Thomas Jefferson University Hospital to automate Qlik application reloads, extracts, report downloads/ distribution etc. Published on pypi as a pip installable package.

Summary

Qlik Engine API Connector lets a user connect to the Qlik Engine API directly using websockets and interact with their local Qlik server effortlessly. Users can automate report extracts with or without Nprinting, distribute pdf/ excel / image reports via email, monitor server usage by hours, check for idle users etc. To explore all the possibilities of the Engine API, go to engine API explorer under dev-hub on your Qlik installation.

Go to the project

Predatory Journal Tracker Chrome Extension

Google Chrome Extension to cross reference journal names from google scholar with Beall's list of predatory journals. This list would give the reader a warning about suspicious papers being passed off as legitimate research by highlighting suspicious papers in red on the google scholar's search page.

Summary

This list is in no way definitive but can give the users a waring sign for a research paper so they can investigate more. Frequently people use flimsy research papers online to peddle pseudo science and fake news. This is specially common in pseudo scientific medicine where people publish articles in "research journals" and make outlandish claims about medicinal benefits of "drinking cow urine".

Go to the project

RateMyProfessors-Banner-TermMaster Integration

Chrome extension to display professor's quality, difficulty scores in Drexel's course search and registration platform. Students can get information about any given course's difficulty or professor's teaching quality right where they need it helping them make better decisions about their course choices.

Summary

Used Jquery, HTML, CSS to create the extension. In process of publishing on the chrome extension store.

Go to the project

STAR Scholars Program Project

This study analyzes gender bias in peer review by analyzing the sentiment of the reviewers comments on the submitted manuscripts.The study attempts to eliminate the quality of papers as an attribute in decision making process and thus compares similar papers to isolate effect of authors’ attributes like gender.

Summary

Publication is at the core of scientific communication and so the decisions made by conferences and journals are non trivial in the effect that they can have on the quality of the work that is being done and as a result the quality of work that gets shared with the world. We use peer read dataset of reviews and analyzing individual reviews that have been presented for manuscripts and analyzing their sentiment.

Go to the project

Automated Echocardiogram Interpretation

Created a pipeline for automated calculation of ejection fraction from DICOM files of electrocardiograms Modified open source tool to gather image segmentation training data Integrated with Hadoop environment

Summary

Automated Echocardiogram Interpretation has potential to achieve great increase in efficiency by automating

Go to the project

CMS AI Challenge

The CMS Artificial Intelligence (AI) Health Outcomes Challenge is an opportunity for innovators to demonstrate how AI tools – such as deep learning and neural networks – can be used to predict unplanned hospital and skilled nursing facility admissions and adverse events.

Summary

Our team got selected in the top 25 for the second round out of 600 applications. Our proposal entitled Using AI to Improve Medicare Population Health, Optimize Ambulatory Scheduling, and Reduce Adverse Events at Hospitals discussed innovative ways to use health insurance claims data to predict health outcomes.

Go to the project

Daily Dilbert Wallpaper

A simple python script to change desktop background with today's Dilbert comic using Python. A happy surprise everyday at work!

Summary

A simple python script to change desktop background with today's Dilbert comic using Python. A happy surprise everyday at work!

Go to the project

CONTACT