May 2020 – Jun 2020 · 1 mo · India

• Built an attachment prediction system for the internal ticketing management system, DESFlow.
• Experimented with various keyword based and ML based algorithms and built a Machine Learning model using the combination of Naïve Bayes and Graham’s algorithm and achieved 90% precision, 59% recall and 90% accuracy.
• Improved the Graham’s algorithm and employed normalized keyword approach on top of the Machine Learning model to improve the recall to 67% and precision to 91%.
• Extended the solution to handle similar use cases like setting the due date and closing the request and also added the support for active learning for model to evolve over time.

May 2019 – Aug 2019 · 3 mos

• Integrated Boost.Units with the existing base coordinate system and restructured all the classes to make them compatible with Boost.Units to provide a robust astronomical coordinate system.
• Created the parser for binary table extension and ASCII table extension for FITS File system.
• Wrote unit tests using Boost Unit Test framework.

May 2018 – Jul 2018 · 2 mos · Kanpur, Uttar Pradesh, India

• Developed a dynamic and scalable web application using LAMP stack from scratch as an initiative to improve themedical system by keeping track of records of patients and their medical history.
• Implemented various functionalities like doctor could add a patient, update its records, refer it to another doctor, etc. and patient could search for doctors and the labs in his area.
• Developed a question-answer platform for the medical system which had the functionality to filter the questions based on illness, search questions based on keywords, notifications for patients if their questions get answered, etc.
• Technologies and languages used: PHP, MYSQL, AJAX, HTML, Javascript, Microsoft Azure(for deployment).