I develop enterprise scale software applications and help maintain them. I love to work on projects that would really create an impact on the world/organization. During my 3.5 years of Software Development career at Morgan Stanley, I have been instrumental in conceptualizing, strategizing and developing complex Financial systems that served exchange of millions of dollars per day. Learning new technologies is essential to my interest. Through out my work experience, I have worked on several programming languages and coding practices. With the recent shift towards Data Science and Machine Learning, my interest towards the field grew and I decided to pursue my Masters at York University with my thesis in Fairness in Machine Learning.
Major Technologies: Java 8/Web Services, Python 3, Perl, AngularJS, Angular Major Projects: 2017-2018: As a part of Collateral Management, I worked on 3 regulatory reporting projects: - Collateral Flag Reporting for EMIR/ISDA - Collateral Reporting for EMIR 2.0 Regulation. - Standard Initial Margin Model (SIMM) as a part of ISDA Regulation 2019: - Onboarded the Repurchase Agreement (REPO) business area in Collateral Management Technology team. - Increased interaction with global users to understand Software Requirements.
Analysts will participate in a 15-week global training program, which includes an orientation to the Firm, intensive technology training, introduction to proprietary technologies, and functional training culminating in a four-week group project. Throughout the Program, Analysts will have the opportunity to build a strong global peer network. The comprehensive skills acquired during our training programs provide the Analysts with the tools necessary for the Associate role.
During the internship, my main tasks were: - Designed Charts and Graphs (In Tableau) to provide a view into the firm's position and risk exposures for OTC, LD and SFT Businesses. - Development of Angular/Tableau UI Dashboard. - Designed REST Services for fetching exposure data.
I am currently a graduate student at York University with the thesis in "Fairness in Machine Learning", My supervisors are Jeff Edmonds and Ruth Urner.
(Google Kaggle Competition) - Preprocessed the Wikipedia talk page comments and trained Bidirectional LSTM with Attention and also fine tuned the BERT model to classify the comments and achieved ROC-AUC of 0.9309 (Top 15%). Report: https://drive.google.com/file/d/1doohTAx2jwNJwQuwWKNxk4FS7kOwFnYj/
View ProjectTrained 3 different models on the Kaggle goodbooks-10k dataset (kaggle.com/zygmunt/goodbooks-10k), and performed analysis to deduce the best model among Neural Nets, K-Nearest Neighbours and Matrix Factorization Recommender Systems. Report: https://drive.google.com/file/d/1doohTAx2jwNJwQuwWKNxk4FS7kOwFnYj/
View ProjectDeveloped REST Services, downstream file feed, JMS for Dowstream client which consumed Collateral Calculation Data according to the new European EMIR regulations. Since this project is proprietary of Morgan Stanley, a live demo is not publicly avaliable.
Developed Python APIs to connect to the in-house Databases (DB2 & Sybase), developed continuous integration frramework using Jenkins & also made a logging framework in JSON format. In addition, I also automated several of my team's daily time consuming routines to save developer's precious time.