Software Engineer with 4.7 years of experience in handling the entire lifecycle of design and development of web-based and stand-alone applications. Pursuing Master's in Computer Science from University of Central Florida. I am interested in the field of Machine Learning and Artificial Intelligence.
Created a CNN to generate a super resolution image for a particular image (png/jpg). The CNN has an input layer, 8 hidden layers and output layer. The input image is convolved to produce a sharp image.
In this project the reviews from Amazon, yelp and imdb are classified to be as good or bad. n-gram model is created to extract features from the text and using Logistic Regression as a binary classifier, classified the reviews into good or bad.
In this project an agent is trained using various types of search algorithms. Using Markov Decision Process, dynamic path planning algorithms, Reinforcement learning, value iteration and q-learning trained an intelligent agent to plan the path to the goal.
Using scikit-learn designed one-vs-one, one-vs-all classifiers and tested with various kernels of SVM(linear, RBF, polynomial, sigmoid). Since the UCI Glass dataset is unbalanced, implemented the weighted one-vs-one SVM classifier.
In this project three types of analysis were performed. First is sending a manual email spam. Second is simulation of a random-scanning worm propagation and local-preference scanning worm propagation. Third, implemented a mutation based fuzzer or fuzz tester for an image.
Development and maintenance of 11 websites of MGMResorts which are used to book hotels, restaurants, cabana and other facilities. Responsible for requirement gathering, design, development of webservices and Adobe Experience Manager(AEM) components using Java.
Development and maintenance of KPMG’s consumer facing websites of 80 countries using AEM and Java.
Designed and developed pharmacy locator tool, registration to a plan, member profile manager tool for the Medicare and Retirement website.
Developed an automation testing tool for website using Cucumber JUnit, comparing the actual and expected behavior and help identify bugs during the development cycle.
Assisting and grading students with Computer logic and design concepts.