Built open-source tool to discover latent information in food science literature using Word2Vec model trained on 3.5 million food science journal article abstracts

Scraped and stored into MongoDB 3.5 million academic journal abstracts from four online databases

Preprocessed data using spaCy and then created custom Docker container to train on AWS SageMaker

Built frontend for serverless web application built on AWS Amplify and AWS Lambda using JavaScript, React, and React-Bootstrap