Bridging the gap between agentic systems and complex customer data

Developed event driven streaming capabilities for time series data at scales of up to 1.3 million datapoints per second

Designed and built tools for sellers on Amazon Fresh/Primenow

· Developed feature that allowed for auto-annotation of images in data sets using Convolutional Neural Networks and Real-time object detection while overcoming pipelining obstacles

· Built a Django based client-server web application utilizing auto-annotation feature that streamlined the process of training Deep Neural Networks on large image data sets by over 800%

· Application is now being used across several teams on ITRI ML related projects such as security camera license plate detection