Projects
Weed ID App
A Flask app to identify plants using characteristics like flower color, height, and geography. The proprietary dataset was curated and sold to large pesticide companies for ML training.
Interviewing.fyi
A platform that transforms interview experiences into career capital. Candidates upload their verified interview experiences via .eml files to get discovered by recruiters.
Oceanographic Research
Published research based on 3 years of oceanographic study. Analyzed biases in upper ocean stratification in CMIP5 climate models to quantify their impact on winter mixed layer depth and ocean-atmosphere interactions in the North Pacific.
Digit Recognizer
An ensemble of 5 TensorFlow.js CNN models trained on the MNIST dataset to recognize handwritten digits. Draw a number and watch the models predict it in real-time. Originally built in 2020.