Cyber Infrastructure-enabled ML

Co-Webmaster @ UCSD JSA■ Planned, designed and implemented both the Firebase/JavaScript backend and HTML/CSS frontend of new features on the JSA website.■ Identified cost and bandwidth limitations, and produced scalable solutions to address them.■ Implemented a login system that allows select members administrative privileges.■ Designed and implemented an image compression/upload system and interface.■ Designed and implemented an interactive image gallery that enables… Show more Co-Webmaster @ UCSD JSA■ Planned, designed and implemented both the Firebase/JavaScript backend and HTML/CSS frontend of new features on the JSA website.■ Identified cost and bandwidth limitations, and produced scalable solutions to address them.■ Implemented a login system that allows select members administrative privileges.■ Designed and implemented an image compression/upload system and interface.■ Designed and implemented an interactive image gallery that enables photo filtering.■ Communicated with executive members to understand their website requirements and proposed ideas, solutions, and improvements.■ Provided regular progress reports and updates regarding the development and implementation of new features.jsa.ucsd.edu Show less

Power Reduction Team @ NXP Semiconductors■ Deployed machine learning models on Amazon Sagemaker utilizing Amazon S3, and communicated results with relevant professionals in the Semiconductor field.■ Employed supervised learning in the form of neural networks, regression, support vector machines, and unsupervised dimensionality reduction techniques such as PCA and TSNE.■ Implemented code to parse data logs for key information, filtering unnecessary fields of data and performed… Show more Power Reduction Team @ NXP Semiconductors■ Deployed machine learning models on Amazon Sagemaker utilizing Amazon S3, and communicated results with relevant professionals in the Semiconductor field.■ Employed supervised learning in the form of neural networks, regression, support vector machines, and unsupervised dimensionality reduction techniques such as PCA and TSNE.■ Implemented code to parse data logs for key information, filtering unnecessary fields of data and performed data processing at the collecting stage.■ Increased efficiency of workflows by refactoring the existing codebase, wrote python scripts that work in conjunction with existing frameworks to run comprehensive simulations on Ansys PowerArtist.■ Significantly decreased data collection time of the team from 2 days to 12 hours by identifying workflow issues, and developed python scripts that worked with .tcl files to maintain dependencies without restarting time consuming simulations. Show less