1. Ritesh Kasat
Email: riteshkasat@gmail.com Phone: (213) 675-3081 Website: riteshkasat.github.io
Links: github.com/riteshkasat | linkedin.com/in/riteshkasat | stackoverflow.com/users/2355376/riteshkasat
Objective
Seeking a Full-time opportunity in the field of Software and Data Engineering starting May 2017.
Education
May 2017 MS, Computer Science (Data Informatics)
University of Southern California, Los Angeles GPA: 3.6
Sep 2014 B.E., Information Science
P.E.S. Institute of Technology, Bangalore, India GPA: 3.9
Technical Skills
Programming Languages: Java, Python, C/C++
Web Technology: HTML, CSS, JavaScript, XML, JSON, AJAX, Bootstrap
Databases: Mongo DB, MySQL, Oracle, PostgreSQL
Data Science: Hadoop, Map Reduce, Scikit-learn, Data Mining, Machine Learning
IDE: Eclipse, IntelliJ Idea, PyCharm, WebStorm
Others: Git, Maven, Gradle, JUnit, Spring, Jenkins, Mockito, Agile, Apache/Tomcat
Work Experience
Jan ’17 - Current Software Co-Op, Trifacta Inc. San Francisco
Developing Python based Web Crawler to crawl all the webpages in Trifacta Wrangler and
take screenshot of each visited page. Developed Mocha based page test cases for Trifacta
Wrangler.
Tech stack Python, JavaScript, Jenkins, Slack, Vagrant, Git, Phabricator, Mocha.
Aug ’16 - December ’16 Student Developer, Perfit Los Angeles
Developed Python based RESTful web-service for Perfit -a 3D modeling software- that auto-
matically drapes clothes on the pre-constructed 3D model of a specific person. Implemented
login and signup API for the website. Integrated google recaptcha authentication for login.
Also responsible for setting up build automation and modularizing codebase using MVC
design pattern. The web-service is hosted on AWS Elastic Beanstalk environment.
Tech stack Python, Flask, RESTful web-service, PyBuilder, PyCharm, MVC, AWS Elastic
Beanstalk.
May ’16 - Aug ’16 Summer Intern, Trifacta Inc. San Francisco
Developed a tool for customer success team that allows automatic import of scripts in the
Trifacta Wrangler. Also developed several JavaScript automations such as automation to
verify HDFS file permissions. Enhanced existing framework by creating new page models
and refactoring existing ones.
Tech stack JavaScript, Postgres, Jenkins, Slack, Vagrant, Sahi, Git, Phabricator.
Aug ’14 - Aug ’15 Software Engineer, Intuit Inc. Bangalore
Worked on the core development of QuickBooks Online, an accounting software used by
millions all over the world. I worked with QuickBooks platform team to develop RESTful
web-service in Java for subscription billing service. I also enhanced a JavaScript widget
for UK Direct Debit payment facility. Test-driven development (TDD) was followed and I
achieved an overall code-coverage of 90% for the new modules
Tech stack Java, JavaScript, MySQL, Netezza, RESTful web-service, Jenkins, IntelliJ Idea, JUnit.
Jan ’14 - June ’14 Software Co-Op, Intuit Inc. Bangalore
Worked with the payments-platform team within Small Business Group of Intuit. During
my co-op, I developed an algorithm in Java for automatic reconciliation of transaction in
QuickBooks. This algorithm reduces 25 hours of manual effort each month to just 30
minutes. I was also responsible for designing exception handling strategy for the platform
reconciliation service. All the code was reviewed, perfected and pushed to the production.
Tech stack Java, JavaScript, Perforce, Git, Jenkins, Eclipse, Gradle, Maven.
2. Academic Projects
• Email Spam Classifier | github.com/riteshkasat/Spam-Classifier | Python, ML, Naive Bayes
This is a Machine Learning Project that distinguishes spams from genuine emails. The model is learnt using Naive
Bayes algorithm. This project also implements an enhancement called add-one smoothing. The performance of the
classifier is determined using metrics such as precision, recall and F-1 score. I achieved an overall F-1 score of 0.98 for
spam and 0.96 for ham.
• Large Scale Text classification of data | Python, Machine Learning, KNN
This project aims at classifying a huge number of web documents into a hierarchy of categories. The large scale of data
called for the use of advanced classification algorithms that could scale. Parallelized k-NN algorithm with the help of
Hadoop Map-Reduce paradigm was implemented to handle space and time complexities of large data. Cosine-vector
similarity was used to compare distances between two points instead of the euclidean distance with absolute distance
in the traditional algorithm.
• Walmart Trip Type Classification | kaggle.com/rkasat | Python, Sklearn, Kaggle
In this project, I worked on a Kaggle challenge by Walmart and secured overall 503rd rank in this competition. The
problem here is to improve the science behind trip type classification, which will help Walmart to refine their segmen-
tation process. I converted data to libsvm format, built models using classification algorithms such as Naive Bayes, K
means, Random Forest and SVM, compared them to find accuracy, precision and recall.
• Movie Recommender System | github.com/riteshkasat/Item-Recommender-System | Data Mining
In this project, I developed a user-based collaborative filtering recommender system for predicting movie ratings by a
user. The ratings of the new movie for the given user were predicted using the similarity of its k nearest neighbors
determined through their Pearson correlation coefficient.
Coursework
Software Engineering Machine Learning Data Mining Data Management Artificial Intelligence
Algorithm Design and Analysis Web Technology Data Structures Compiler Design Operating Systems
Computer Networks Computer Networks and Security Natural Language Processing OOP with Java
Hackathons
October ’16 Money2020 Hackathon, The Venetian, Las Vegas
Developed an app to order food using Amazon’s intelligent personal assistant Alexa. Using
voice interaction, the user can order food under 1 minute. The payment for the ordered
items is done using Visa Checkout API. A confirmation is also sent to the user using Twilio
API.
Tech stack Amazon Echo, JavaScript, MySQL, Twilio API, VISA API
April ’15 Data Science Hackathon, USC, Los Angeles
Developed an app that could automatically detect interesting topics from the tweets. Topic
detection was done using Latent Dirichilet Allocation - an unsupervised Learning
algorithm. The topics were visualized using a python package called pyLDAvis.
Tech stack Python, Machine Learning, Clustering
Achievements
• Awarded with the title of Amul Vidya Shree for outstanding academic performance in CBSE class 10th Board
examinations and securing first rank in the Guna District of Madhya Pradesh.
• Awarded with scholarship and certificate of First Class with Distinction for all the semesters in the undergrad
course.
• Awarded with silver medal in National Mathematics Olympiad for securing 96 percent marks.