KASI PERUMAL SUNDARAJ

adldqm@r.postjobfree.com 407-***-**** https://www.linkedin.com/in/kasiperumal/ ACADEMIC QUALIFICATIONS

University of Central Florida, Orlando, Florida December 2018 – May 2020 Master of Science in Data Analytics GPA–3.65

Anna University, India August 2012 - April 2016

Bachelor of Engineering in Computer Science & Engineering GPA – 3.5 WORK EXPERIENCE

MACHINE LEARNING DEVELOPER STREAMLYNE January 2020 – Present

• Working on Unsupervised Machine learning Algorithms and Natural Language Processing in the part of application development. This includes Data preprocessing, Text parsing and Normalization over unstructured data.

• Also implementing various python packages, modules, and libraries that support business in decision making and solve real-time problems. This functionality uses NumPy, TensorFlow, scikit-learn, and other data analytics tools. As a part of the application development JIRA, Bitbucket, Git, and AWS Sage maker, Azure, GCP were also used. GRADUATE RESEARCH ASSISTANT UNIVERSITY OF CENTRAL FLORIDA September 2019 – Dec 2019

• Worked on Deep Learning and Natural Language processing using Recurrent Neural Network. Initiated a project to work on improvising the Chatbot algorithm using the RNN network.

• The project aims to answer user queries which include nested intents, and this method allows the use of standard constituency parsing models to outperform sequence-to-sequence baselines on the sample datasets. PROGRAMMER ANALYST COGNIZANT TECHNOLOGY SOLUTIONS May 2016 – Dec 2018

• Worked as a Programmer Analyst in supply chain management for our client “The Clorox Company”. Experience in handling large data processing including structured and unstructured data

• Handled various Data security and Data Migration tasks for Client Applications. Actively involved in writing automation suites for application processing semi-structured data through servers which in turn saved 1.4 Million US Dollars for our client.

TECHNICAL SKILLS

Programming Languages Python, R, Scala, JAVA, D3.js Packages Scikit-Learn, NumPy, Pandas, TensorFlow, Keras, OpenCV Database System MySQL, Oracle, HDFS, Postgres database Big Data Tool Hadoop, Apache Spark, PySpark, AWS, GCP, UpCloud. IDE Anaconda, IntelliJ, Eclipse, Lotus Notes, Jupiter Notebook, and SAP web IDE Certifications Python Developer, Machine Learning, Deep Learning, TensorFlow and Big Data Research Paper Reviews Deep Learning in Insurance Industry, MapReduce System. TECHNICAL PROJECTS

Project 1: COVID-19 White House Data Challenge– Python (NLP, AWS SageMaker, PostgresSQL) The goal is to support the government to fight against COVID-19 by choosing top researcher papers related to various tasks like providing vaccines, safety measures. This project uses Natural Language Tool Kit (NLTK) and AWS SageMaker to build the neural topic model over thousands of research papers provided by the White House. Project 2: Rock Paper Scissors Detection using Keras VGG19 – Python (TensorFlow, Keras, OpenCV, NumPy) The VGG19 model consists of 20 Million pre-trained parameters and this model was implemented over 4 thousand images of hand (Rock/Paper/Scissors), with 10 epoch and 64 batches the model was able to provide 91.5% accuracy while testing. Project 3: Movie Recommendation System using ALS – PySpark (Spark, Python, K-Means Clustering) The goal of this project is to predict the movie rating system. If a user reviewed some movies, the system would predict ratings for other movies that were not reviewed. K-Means clustering is used to recommend movies based on the ratings. Project 4: Online Product Detection using Convolutional Neural Network – Python (TensorFlow, Keras) The convolutional neural network was used to classify the product based on its 3-Dimensional image. Over 1 Million images were used to train the convolutional model. The project is completely built on Google TPU with approximately 3 million parameters for training data and which produced an accuracy rate of 91.4% over its test data