▪Implemented a scalable ETL pipeline using Python, Spark, and Apache Airflow to fetch sports data from APIs and store it in PostgreSQL database, resulting in a 30% reduction in data errors.▪Built a Natural Language Generation (NLG) system using directed acyclic graphs (DAGs) to transform various API data feeds into auto generated, real-time game recaps for the English Premier League.▪Developed an accurate sports prediction model using the ELO ratings algorithm, achieving an 83% accuracy in forecasting win probabilities for teams across multiple leagues, including EPL, NBA, and NFL.▪Created comprehensive test cases using Python’s unittest framework to achieve 100% code coverageensuring reliability of critical application components.

▪Led a team of 3 developers in engineering a Python application using Flask and RabbitMQ, that streamlined the onboarding process of new customers, reducing onboarding errors by 50%.▪Architected and executed a highly scalable database schema on PostgreSQL for a medical platform called ROCC, reducing query times by 40%.▪Designed and developed secure REST APIs using Java SpringBoot and Python, connecting the ROCC user interface with backend infrastructure; reducing API response time by 25%.▪Deployed over 5 microservices on Kubernetes using Docker containers, ensuring high availability.▪Wrote fast and efficient queries using GraphQL API, reducing the application load time by 24% and patient connect time from 8 seconds to 3 seconds.▪Implemented the entire frontend application for ROCC using React.JS, Typescript, Redux and Node.JS,enabling real-time communication between radiologists and technologists.▪Deployed a Flask application on an AWS EC2 instance, ensuring seamless access to the application.Configured Nginx as a reverse proxy to enhance performance, scalability, and security.▪Collaborated with cross-functional teams to troubleshoot issues, resulting in timely resolution of defects and improved system reliability.

▪Collaborated with the development team to build a Recommender System for educational modules,using Matrix Factorization to predict ratings for each user and item, enabling customers to discover relevant and engaging content.