Shannon Reuben Serrao

Blacksburg, VA 540-***-**** adntap@r.postjobfree.com github.com/shannonserrao F1-OPT Technical Skills

DATA SCIENCE: Pandas NumPy Scikit-Learn Signal Processing Image Processing SciPy PROGRAMMING LANGUAGES: Python SQL C++/C MATLAB Fortran Bash Mathematica COBOL MACHINE LEARNING: Time-Series modelling(ARIMA) Regression NLP (GloVe, FastText) Classification Clustering Deep Learning (Pytorch) Random Forest K-Means TensorFlow, Keras Distributed Computing Apache Spark SVM AWS Gradient Descent Naive Bayes Neural Networks (CNN, RNN, LSTM) Price prediction Big Data DATA VISUALIZATION WEB: Matplotlib Bokeh Altair Seaborn Flask HTML MATHEMATICS AND STATISITICS: Hypothesis Testing Statistical Modeling Monte Carlo Methods Linear Algebra DOMAIN KNOWLEDGE: Finance Epidemiology Churn Analysis Projects and Achievements

LEASETOME WEBAPP RECOMMENDER SYSTEM github.com/shannonserrao/L2M_Capstone

• Data Engineering using Graph QL to pull housing data from third party APIs; Built a recommender/visualization webapp targeted at commercial retail space customers for LeaseToMe. TEXT BASED VISUAL, QUESTION ANSWERING IN IMAGES github.com/shannonserrao/textVQA_challenge

• Developed a vanilla VQA model with a multidisciplinary team, improving accuracy by 38%; maximized file storage by performing data cleaning while engineering and refining OCR characters in image and conversion to fastText and GloVe embedding vectors; used CNN(Pytorch) to implement OCR attention to answer image-based questions AN ANALYSIS OF STACK OVERFLOW SURVEY github.com/shannonserrao/Stack_Exchange

• Improved Stack Overflow customer satisfaction with multi-class classification, regression and clustering based analysis COVID NETWORK MODELING github.com/shannonserrao/COVID_SEIR

• Created a small-world lattice model to simulate COVID outbreak in a community as part of the pandemic prevention program at Virginia Tech; predicted spread by restructuring real-world data and recommended preventive measures; re- structured real-world graph data to predict/test preventative measures IMAGE ANALYSIS OF ECOLOGICAL MODELS github.com/shannonserrao/May-Leonard_lattice_Simulation

• Built a Monte-Carlo model to analyze May-Leonard ecology; utilized spiral image data to determine spiral stabilization criterion; detected temporal and spatial correlation signal and pattern recognition of spiral structures in image data using Pandas and SciPy; mentored three undergraduate students for three years AWARDS: WAN-ZIA SCHOLARSHIP FOR EXCELLENCE IN SCHOLARSHIP AND ACTIVE ENGAGEMENT OF RESEARCH Work Experience

HSBC-GLT PUNE June 2008 – August 2009

Software Engineer

• Processed vast amounts of financial ledger information to write COBOL codes for Financial Straight Through Processing on mainframe computers

Education

Ph.D in Physics Virginia PolyTech Institute & State University December 2020 MS in Physics Indian Institute of Technology Roorkee June 2013 BE in Electronics and Telecommunication Mumbai University (Don Bosco Institute of Technology) June 2008