Consumer Group Team
•Built a model to generate an ontology (hierarchy) out of TurboTax free-text data
•Used Google’s word2vec word embeddings for text processing and hierarchical clustering to generate clusters for ontology
•Built end to end interface using React.js, Flask, and ElasticSearch to query and interact with ontology,
•AWS ElasticBeanstalk, ElasticSearch cloud, and EC2 instances used for system deployment