Designed and implemented an ML dashboard from scratch using JavaScript, GraphQL and S3 in Retools to streamline the analysis of datasets, checkpoint inferences and metrics

Built a GraphQL API by setting up an EC2 instance with a Flask endpoint and fetching training run data from DocumentDB for an efficient command-line to web tooling migration

Built out image and text embedding features with PHP and Python for 3D videos in VR using computer vision and NLP services

Improved ML infrastructure with C++ by adding a configuration based universal value model to save ~6 hours from model experimentation and deployment

Added an image fallback mechanism with C++ to reduce each request’s image configuration size by 6.5x and to preserve image consistency across platforms