Data Engineer/Scientist

Suresh Kasipandy

**** * ******* ** *** ***, San Mateo, CA 94403 323-***-**** ******.*********@*****.*** Objective

To secure a full-time employment or internship position at a company to hone my skills as a Software/Data Engineer as well as gain valuable knowledge and experience in the field of programming and software industry. Education

BACHELOR OF SCIENCE (HONS) COMPUTER ENGINEERING CALEDONIAN COLLEGE OF ENGINEERING

Learned computer networking, hardware, security and coding (C and C MASTER OF SCIENCE DATA INFORMATICS UNIVERSITY OF SOUTHERN CALIFORNIA

Gained proficiency in Python and Big Data related fields like Machine Learning, Data Mining and MapReduce (Hadoop).

Gained a good understanding of Natural Language Processing (NLP) techniques and the basics of Artificial Intelligence. Experience

WEB DEVELOPER DIGITAL INTELLIGENCE JANUARY 2014 – AUGUST 2014

Helped design wireframes, mock-ups and sitemaps for websites and develop websites using HTML and JavaScript.

Conducted research on big data as well as upgrading systems to accommodate big data related technologies.

Learned how to manage data in RDBMS using SQL.

Projects

WALMART TRIP TYPE CLASSIFICATION

Created model using Decision Tree Algorithm and toolkits such as Scikit-learn and Pandas to predict trip types. SENTIMENTAL ANALYSIS OF SONGS USING TEXT CLASSIFICATION ON SONG LYRICS

Implemented a variety of classifier algorithms (Multi-layer perceptron, Adaptive boosting classifier, Maximum entropy classifier) with hyperparameter optimization.

Data was preprocessed by NLP techniques (tokenization, special character removal, whitespace trimming, lowercase) and tested with other modifications (stop word removal, n-grams and stemming) for best accuracy. SANTANDER CUSTOMER SATISFACTION PREDICTION

Created model for predicting customer satisfaction for unlabeled customers using combination of Random Forest and Extreme Gradient Boosting.

SEQUENCE LABELING

By using the CRFsuite toolkit, a baseline set was creating using provided labeled data which was then used to label utterances in unlabeled conversations with dialogue acts. SPAM FILTER

Used Naïve Bayes and Perceptron (standard and averaged) algorithms to classify emails as spam or ham.

Contact this candidate