Savar Sharma

Developed intuitive and accessible web components using TypeScript in AngularJS and HTML Native Web Components. Wrote automation tests in ROBOT framework in Python and TypeScript. Maintained CI pipeline on Jenkins to facilitate Acceptance Test Driven Development.

Developed web application in Reach using Typescript, for end-to-end Sanitation Management in the district of Kurukshetra, India. Developed highly accessible and robust React based web application supporting touch screen interfacing to conduct driver license exam.

Development of forecast models for agricultural systems covering agricultural, horticulture, veterinary sciences, dairy, poultry, fishers, etc. Developing various linear and nonlinear time series modeling and forecasting in Agriculture.

Working on different aspects of a project from designing user interfaces to implement server side logic and database management. Practice prototyping in order to quickly prototype and develop minimum viable products (MVPs) to test ideas and gather user feedback.

Developed a comprehensive active learning model which aims to improve the accuracy of other Machine Learning models with a lot less training examples, using Long Short-Term (LSTM) method. Improved semantic segmentation of satellite imagery using Convolutional Neural Networks; employed un-sampling and masked R-CNN.