Summer Intern - Full stack developer
2016 (3 months)
As a team of four, designed and developed a web application for getting personalized news feed based on topic modeling and gamification of news content (point system) - AngularJS,...more Angular material, Django REST framework, Elastic Search.
The app is hosted on medialistener.apollobit.com
Implemented a recommendation engine which suggests topics to users based on history and added to user preference page in which user can also add topics, locations, publishers to their interests- Grails, Python.
Bangalore - Software Engineer I & II
2012 - 2015 (about 3 years)
Developed Automation Framework and automated scripts for verification of Diameter protocol messages and multiple traffics like TCP/UDP/SIP with deep packet inspection of Firewall/NAT/ALG...more - TCL/Expect
Developed Web Scheduler application which is used by multiple teams for scheduling test suites, capturing test results, mapping failed test cases to existing ticket by analyzing error logs– JSP, Java
Feature development of Priority Indicator for restoring Voice over LTE bearers during migration- C
Web application for reading personalized news articles
Web developer, Data Engineer, AngularJS, Python/Django, Elastic Search · Getting personalized news articles, engage users by assigning scores to user… · More based on shares, time spent, game points after reading articles.
We followed Agile technology. We used to discuss features that need to be developed at the start of each sprint. There was a demo for each week. I developed a feature for fetching/displaying the news articles in list/grid view and retrieve/display statistics of articles. Hot topics/suggested topics based on preferences are displayed.
Used AngularJS, Angular material for frontend, Django REST framework for creating API endpoints, Elastic Search for indexing data
Created dataset and recommendation engine which suggest topics to users based on history and added to user preference page
Web application with following features 1) User preference for adding locations/topics/publishers of interest 2) Dashboard for viewing scores for each section which is calculated based on shared articles/time spent/game score after reading each article/read article per topic 3) Suggest recommended topics to user based on history/topics of interest 4) My articles section retrieving articles in list/grid view. 5) Search/Browse topics/keywords/articles 6) Compare keywords 7) Game content for each article
Classification of data to eliminate vulnerabilities(identification of program flow from input) of various timing side-channel attacks. Execution time was… · More clustered. Various machine learning techniques like K-means, Decision tree, Neural Nets, Adaboost, and SVM were tried for classification. Linear regression was used to find program execution time for a particular input
game results were predicted by classification of tweets, HMM
Developed results prediction for NBA with 65-70% accuracy using live twitter feeds. Tweets were classified, map-reduce algorithm was applied to reduce tweets… · More and ran sentiment analysis using Naïve Bayes classifier; Predicted the winner based on the support for a team/players and previous results using Hidden Markov Model
What I Do
I am Graduate student at the University of Colorado Boulder specializing in the field of Data science.