Worked as backend developer in Real Time Payments (RTP) team. RTP builds next generation payment network connecting individuals and businesses with banks to enable payment transfers in real time.- Worked on Message Level Encryption feature to provide safe and secure transfer of sensitive information supporting both payload encryption and digital signature verification.- Designed and developed highly resilient events driven notification pipeline. Standardised various event related formats using cloud events spec for effective processing and maintenance. - Implemented health check service between participants and RTP system by exchanging sign on/off and echo messages.- Contributed to multiple cron jobs to improve overall fault tolerance with a generic retry and timeout mechanisms. - Automated functional tests to ease features verification in multiple environments.- Actively involved in collaboration with various teams with a focus on improving performance metrics mainly decreasing End to end transaction latency and increasing TPS. Developed automated scripts for performance measure of async transaction calls.

My team has the mandate to "Drive Linux system performance". Primarily I spent my time on:▪ Worked on Hibernation feature to reduce device boot time by 200% by avoiding user space initialization and complete shutdown gives significant power savings.▪ Proposed new solutions for boot time improvement by resolving the issues of running multiple subsystems in parallel.▪ Received QualStar award for outstanding contribution to a proprietary project by shedding thousands of lines of code.

Topic: Parts-of-Speech tagging of resource poor languages. ▪ Proposed a novel Dynamic Classifier Selection approach for Parts-of-Speech tagging of resource poor languages.▪ Executed experiments on five Indian languages and the enhanced results outperformed base classifiers. On an average, POS tagging accuracy improved by 10% for all the experimented languages.▪ Proposed an efficient heuristic for finding nearest neighbors in KNN classification algorithm. Our experiments conducted on the datasets taken from UCI repository shows a mean improvement of 5% accuracy compared to standard KNN.

Worked with the team in organizing Monsoon'15 placement drive.

Appointed twice for Introduction to Databases course(Monsoon’14 & ’15).Worked on setting the course structure, conducted tutorials, quizzes and evaluated examinations.