DashRide is a sophisticated web application designed to provide seamless ride booking and ride request services and is developed to ensure reliability, scalability, and user-friendly interaction.
Developed custom ASCII encoding and decoding routines for both integer and floating-point types in Apache Arrow. The solution builds upon the design principles of the existing PlainEncoder/Decoder, extending the functionality to provide ASCII representations with an emphasis on readability and precision.
The COLORX is an Android medication reminder app tailored for color-blind users. By integrating REACT NATIVE, PYTHON, FLASK, and MYSQL. ColoRx creates a robust backend infrastructure. The app's key features included adaptive color schemes, family notifications, and medication history tracking, resulting in a significant 40% increase in medication adherence among its users.
MAK.AI is a Full Stack ML Web & Mobile Application for predicting corn plant diseases. By utilizing DJANGO, POSTGRESQL, REACTJS, REACT NATIVE, and MATLAB, the project significantly reduced crop yield loss by 40%. Optimized ML models tailored for mobile devices ensured efficient performance, empowering farmers with timely insights to safeguard their crops and boost agricultural productivity.
The project evaluated testing and debugging tools using real-world bug datasets like Defects4J, Bears, BugSwarm, and QuixBugs. Utilizing technologies such as JAVA, RANDOOP, EVOSUITE, CLOVER, PYTHON, and LEVENSHTEIN, the team generated test suites and conducted code coverage analysis. Metrics like Cyclomatic Complexity Change (CC) and Levenshtein distance (LD) were assessed, aiding in bug localization based on suspiciousness scores. This comprehensive analysis provided insights into effective bug detection and fixing methodologies.
BEATMETRICS is a music discovery web app with ML-based recommendations. Using REACT.JS, NODE.JS, EXPRESS.JS, MYSQL, FLASK, GCP, and TENSORFLOW, the team ensured robust functionality. Flask powered the backend API, while strategic indexing and query caching slashed query response times by 60%. Python optimizations enabled a fastAPI to efficiently serve over 1000 users per second.