E-commerce

Developed a chrome extension for mobile web browsers that extract on-site data, mapping to product information from third-party sites. That data was processed using Random Forrest and NLP classification models to output lower price product comparisons. This new approach leverages a real-time human-in-the-loop system to maximize accuracy.

‍Developed machine learning real-time dynamic algorithm using random forest technique based on customer profile and history to predict most impactful product placement adjacent to primary product being searched.

‍Built a classification and recommendation engine based on a vector clustering system.

‍KUNGFU.AI built a recommendation model trained on a taste graph that mapped vectors of taste onto properties of artists and open-ended questionnaires which lead to an increase in conversions

We trained models to identify a subset of rare coins with over 90% accuracy with a roadmap to higher accuracy and multiple business use cases for deployment.