- Engineered robust forecasting models (XGBoost, LightGBM, and traditional time series models) to accurately predict cloud spend across four key segments: Entertainment/ESPN, Studios, Experiences, and Corporate- Developed key components of an AI platform that facilitates seamless integration of document ingestion, retrieval systems, generative models, and model evaluation

- Developed offer scoring system to improve user experience and reduce turnaround time for sneaker trades- Predicted likelihood of trades being accepted using Logistic Regression, XGBoost, Decision Tree, and Neural Network models - Conducted SMOTE to handle class imbalance and evaluated model performance with accuracy, precision, recall, and F1-score

- Created a subscription-based product for retail traders presenting real-time social media statistics for equities- Leveraged NLP and LLMs with prompt engineering to develop a custom sentiment analysis model with over 85% accuracy- Collaborated with cross-functional teams to develop both frontend and backend systems and built an efficient data pipeline to deliver real-time statistics linking media trends to equity performance

- Utilized Tableau to create data visualizations and interactive budget dashboards resulting in $3 million annual savings- Extracted and presented custom performance metrics for various business units to inform financial decisions