-> Developing exciting AI technology
-> Designed and wrote quality assurance metrics, increased clarity of output by 10 %
-> Automated an Image selection process for BigSur's IP, used CLIP model
Along with 4 other students and the teaching professor, reconstructing the intro to CS course at Caltech. Transitioning from a skills course to teaching computing as literacy course
Documentation, testing, and software architecture; 5-week long-term project, testing, and extensibility then using POSIX and APIs to develop large software deliverables.
Constructed a classification model which identified motility-based biosignatures in water. A strong detector of microbial life for JPL scientists to deploy in space.
Developed an unsupervised Hidden Markov model employing the Viterbi algorithm, enabling connections between word-phrases, and subsequently trained this model to produce unique Shakespearean sonnets.
Trained an image classification network on the Chexpert chest xray set. Optimized the model’s performance through rigorous hyperparameter tuning, achieving superior diagnostic results.
Constructed an enhanced dynamic memory allocator leveraging an explicit free list, incorporating coalescing and block-splitting methodologies to improve both the allocator's efficiency and performance speed.
Built a comprehensive full-stack website leveraging Django,Python, HTML, CSS, and JavaScript. Optimized website performance using modern web development best practices, including code minification, bundling, and browser caching techniques.