Implemented big data solutions and created in-house Hadoop clusters for the data collected by IoT devices and M2M (Machine to Machine) application.
Analyzed data coming from various IoT devices in public distribution system environment, using MapReduce and Spark.
Developed web application to show the yearly changing trends in public distribution systems in India.
To have an accurate time stamp in the data collected by the M2M application, designed PTP (Precision time Protocol) & NTP (Network Time Protocol) infrastructure for time synchronization (in milliseconds) in the PAN India (29 Site Locations) application and database servers running the application. One PTP server and four NTP servers at central location were designed with each remote location having one local NTP Server.
Implemented central logging of system-logs and command-logs in the PAN-India setup of M2M application and database server.
Disaster Recovery planning and bash scripting to automate the DR process for the M2M application without loss of data. These scripts were integrated in the a GUI and the DR declaration and DR take back processes could be done in just few clicks.
Automated the customization of OS to make it compatible with an inhouse application of CDoT. This customization pertains to various types of users with different priviledges in different directories. The application runs at various remote locations. This automation helped to install the customized OS from a central location.
Implemented kNN (k-Nearest Neighbor) algorithm in MATLAB to classify degree of skin burns in a dataset of almost 300 images. Also implemented various burn models - Weaver & Stoll, Fugitt, Takata, Wu, Henriques, Diller and Klutke, Mehta & Wong and Torvi & Dale and Henriques Damage Integral in MATLAB to find the variation of injury parameter wrt time.
Developed automated test scripts in python to setup test environment, fetch test data from database and generate the reports as per business requirements & ensure delivery of quality software applications.