Sign in

Software Engineer Active Directory

Harrison, NJ
March 13, 2018

Contact this candidate



Rutgers Business School Newark, USA

Masters in Quantitative Finance Aug 2016 – May 2018 Courses: Statistics and Machine Learning, Financial Time Series, Econometrics, Object Oriented Programming. University of Mumbai Mumbai, India

Bachelor of Engineering in Electronics and Telecommunication Aug 2009 – Jun 2013 WORK EXPERIENCE

OptionsPlay New York City, USA

Data Scientist Intern Jan 2018 – Present

Built an exploratory system to analyze various machine learning classification algorithms such as Random Forest and ADABoost

Explored various feature engineering and selection techniques using data visualization on Tableau

Implemented Long Short Term Memory (LSTM) algorithm with an accuracy of 64% to predict one month standard deviation stock movement hence increasing the returns of a covered call investment strategy

Classified stocks using k-mean clustering into groups of optimum delta to trade for covered call strategy BlackRock/ Rutgers Business School Newark, New York City, USA Machine Learning Researcher Jun 2017 – Sep 2017

Worked with the Head of Liquidity Research at BlackRock on using machine learning to develop a solution to predict mortgage prepayment.

Identified significant features for mortgage prepayment, gathered data from various sources and built structured dataset by cleaning, processing and organizing raw data (23million entries) to back-test the model performance. Explored Synthetic Minority Over-sampling Technique (SMOTE), Random Over-Sampling Examples (ROSE) and Undersampling to handle imbalanced data

Applied various machine learning techniques such as Deep Learning, Long Short Term Memory (LSTM) and Gated Recurrent Unit

(GRU) using Keras to predict mortgage prepayment. Analyzed and implemented various regularization techniques such Elastic Net regularizer, Ridge regularizer, LASSO regularizer and Dropout to reduce model overfitting and improve generalization

Achieved accuracy ranging from 76% to 81% and evaluated results with those of Logistic regression based on f1-score and ROC curve (AUC score: 0.65)

KPMG Mumbai, India

Cyber Security Analyst Jun 2015 – Jun 2016

Implemented Random Forest to detect phishing websites for various companies in telecom domain with an accuracy of 78% to 84% and thus reducing the user information leak and fraudulent activities due to phishing attacks

Took initiative and built a model to extract tweets to identity for malicious users and tweets causing reputational adversity to the client by compromising client’s internal information on the social media platform. Applied unigram and bigram approach to identify tweets pertaining to information leak and performed sentiment analysis on tweets by implementing SVM (Support Vector Machines) with an accuracy of 72%

Built a dashboard on Tableau to aggregate and display the top phishing sites, leaked documents, sentiments with the graphs displaying monthly patterns

Accenture Services Private Limited Bangalore, India Associate Software Engineer Nov 2013 – Dec 2014

Implemented various automation frameworks using Python enhancing the operational efficiency by 40%. Supported IT infrastructure using technologies which include Active Directory, SCCM, SCOM PROJECTS

Facial Expression Recognition (ConvNets and RNN) Mar 2017 – Apr 2017 Explored Recurrent Neural Network as a possible solution for image processing. Implemented facial expression recognition system using ConvNets and LSTM to detect discreet facial expressions and achieved an accuracy of 62%. Forecasting of Crude Oil Prices Mar 2017 – Apr 2017 Fitted a Vector Autoregression model to predict crude oil prices, checked for co-integration using Johansen test and then fitted an error correction model to remove co-integration, forecasted oil prices using restricted VAR model. Pricing Fixed Income Securities Using Clustering Nov 2016 – Dec 2016 Developed a solution to create next generation pricing model using machine learning technique. Analyzed different ways to measure similarity between fixed income securities and applied multiple algorithms including K-mean, Hierarchical Cluster to classify and price similar securities.


Languages and Tools: Python (TensorFlow, Keras, Numpy, Scipy, SciKit-learn, NLTK, Pandas, Matplotlib), C++, R, SQL, MongoDB, Tableau.

Contact this candidate