Projects

A Python-based CLI assistant that helps you dominate your fantasy football draft using local AI. Combines VORP-based player rankings with real-time news analysis via Ollama to suggest optimal picks.

An open-source platform to explore and build smarter prior authorization rules. Features a provider-focused web app that showcases crowd-sourced rule definitions, an AI assistant, and a smart form builder—all optimized for transparency and automation.

Led GenAI initiative in prior authorization, automating urgent request detection with high accuracy and adding AI clinical summarization capabilities.

Built predictive model for medication non-adherence, published in 2023 ISPOR. Led cross-functional outreach on 1,500 Medicare members identified by the model.

Spearheaded development of in-house GenAI to answer company policy questions, streamlining internal knowledgebase access and improving efficiency.

Implemented GenAI automation for customer service and MCHAPS survey analysis, cutting turnaround from one week to minutes.

Managed annual wellness visit outreach model, coordinating delivery of messaging to 30,000+ members to improve preventive care engagement.

Devised risk driver model to interpret 1,000+ parameters of internal risk model, providing insights for healthcare risk assessment.

Transformed legacy sklearn models to PySpark pipelines, optimizing for Microsoft Azure cloud infrastructure.

Created web feedback analysis tool using topic modeling and sentiment analysis to summarize thousands of customer surveys, presented results in PowerBI dashboard.

Implemented LSTM model to predict motion in video; co-authored paper published in IJCNN 2018.