Rohan

Software Developer

Amdocs•  July 2019 - May 2022

Developed and implemented a cutting-edge Token Management System using Microservices, REST API, Spring, and Java resulting in a 75% reduction in potential fraud of Employee Preferred Pricing options at client’s store locations. Collaborated with cross-functional teams to deliver a portal for self-activation of electronic Sims via seamless interaction between Amdocs Ensemble API and third-party interfaces resulting in a 50% reduction in customer activation time. Efficiently resolved internal and external change requests to ensure production stability, resulting in cost savings of over $10 million for the client. Imposed ardent measures that reduced the occurrence of future change requests by 30%.

Data Scientist

Ten•ax Innovation Lab•  December 2018 - March 2019

Spearheaded SENAUTO, an innovative autonomous review and sentiment analysis system for an FMCG client, incorporating radical techniques such as Facial Microexpressions, Speech Sentiment, and Intonation Analysis. Implemented a state-of-the-art Facial Micro-expression recognition algorithm in Python utilizing the FERPlus dataset, Keras framework, and optimization with Microsoft’s CNTK architecture, achieving an impressive accuracy rate of 86%. Redesigned Google’s Speech-to-text NLP API for Human Speech Sentiment and Intonation Analysis, deployed it on the cloud using Docker, GCP, and visualized the results with interactive graphs using Plotly, Matplotlib, and Seaborn.

Java Developer

Dotsquares•  June 2017 - July 2018

Orchestrated the development of a dynamic website using HTML, CSS, and Javascript, resulting in a 30% increase in user engagement and a 20% decrease in bounce rates, attributable to the implementation of a categorical area-wise shop search system using Java, JSP, JDBC, and Servlets. Engineered an intuitive admin panel facilitating a 50% reduction in the time required for administrators to manage shop data, achieved through seamless addition, removal, and editing functionalities. This enhancement led to a 25% improvement in website content accuracy and relevancy.