Portfolio Overview

H2 Health – Physical Therapy Study

2025-2026

The most technically comprehensive project to date, this processed patient speech from 2,000+ physical therapy patients using a fully agentic AI bot built on GCP and Vertex AI with Google Playbooks and Gemini LLM calls. Audio was transcribed and analyzed for personality traits via a custom classifier suite, with multiple models trained and evaluated. Clustering and dimensionality reduction produced actionable patient segments, all visualized for the clinical team.

Agentic AI, ML Model Design and Implementation, Classifier Development, GCP / Vertex AI, Google Playbooks, Transcription, Pipeline building, Python, NLP, Healthcare, Data Visualization

Mayo Clinic – Transplant Patient Study

2026

Agentic chatbot study exploring transplant patient-reported outcomes at Mayo Clinic. The AI bot conducted structured patient interviews about transplant success, with responses transcribed and processed through an NLP pipeline to extract personality features. Clustering and dimensionality reduction identified meaningful patient profiles, with visualizations delivered to the research team. Built and deployed on GCP with Vertex AI, with custom Google Functions and APIs.

Agentic AI, GCP / Vertex AI, Transcription, Pipeline building, Python, NLP, Healthcare, Data Visualization

Rush University – CPAP Adherence Study

2025–2026

Large-scale text-based NLP study examining personality traits across 1,000+ patients enrolled in a CPAP compliance research program. The pipeline performed automated feature extraction, clustering, and dimensionality reduction to identify personality-driven adherence patterns, with results visualized and reported to the research team. Deployed on Azure.

NLP, Text Analysis, Pipeline building, Python, Pandas, Azure, Healthcare, Data Visualization, Surveys and Reporting

Dallas Renal Group – Dialysis IVR and Study

2025

Patient feedback study for a Dallas-based dialysis provider, collecting and analyzing IVR-based interviews from 200+ patients. Audio responses were transcribed and processed through an NLP pipeline to extract personality traits, with clustering and dimensionality reduction surfacing distinct patient profiles. Insights were visualized and delivered to the clinical team. Built on GCP with Vertex AI, Dialogflow, and custom Cloud Functions and APIs.

Transcription, IVR Design, GCP / Vertex AI, Dialogflow, Azure, Pipeline building, Python, NLP, Healthcare, Data Visualization

Cleveland Clinic – Hair Restoration Study

2025

Personality-focused analysis of post-procedure patient feedback for Cleveland Clinic's hair restoration program. Audio interviews from 200+ patients were transcribed and fed into an NLP pipeline that scored individuals on key traits, with clustering and dimensionality reduction revealing distinct patient segments. Findings were visualized and summarized for the clinical research team. Deployed on Azure.

Transcription, Data Transformation, Pipeline building, Python, Pandas, NLP, Healthcare, Azure, Data Visualization

Loma Linda University – Bilingual Care Study

2025

Cross-language personality analysis study examining care outcomes for Spanish- and English-speaking patients, with a cohort of 60 participants. The project analyzed existing personality scoring models for bilingual use, exploring whether trait profiles and clinical insights held consistently across both languages. Findings contributed to more equitable, culturally informed care strategies.

Cross-language NLP, Data Transformation, Pipeline building, Python, Pandas, Azure, Healthcare, Surveys and Reporting

Attitudes on Aging Study – McMaster University

2024

Focused on UK and Canadian seniors, this project analyzed elderly individuals' attitudes toward aging, identifying key personality-based attitude clusters and demographic insights. Python-driven NLP and statistical methods highlighted meaningful differences across traits, with findings presented in a detailed report. This survey also included a thorough demographics analysis.

Demographics, Data Visualization, Python, Pandas, NLP, Surveys and Reporting

ESG Weight loss Study - TrueYou

2024

Study exploring the relationship between personality traits and patient satisfaction with endoscopic weight loss treatments. This involved data transformation, of programmatically transcribing patient interviews to get text data, then using NLP to score patients on different personality metrics. Analysis revealed outcomes correlated primarily with initial weight rather than personality, guiding patient expectations and treatment approaches.

Transcription, Data Transformation, Pipeline building, Python, Pandas, NLP, Healthcare, Weight loss

Smoking Cessation Study - J&J

2023-2024

This study analyzed over 6,000 participants' traits to identify effective quitting strategies. Using NLP, clustering, dimensionality reduction, and building a custom ML model, the study recommended personalized cessation methods based on personality profiles. Data visualizations were generated, and the insights were presented in a PowerPoint summary.

ML Model Design and Implementation, Applied AI, Pipeline building, Python, Pandas, NLP, Healthcare

Low Calorie Diet Analysis - NHS

2023-2024

This analysis correlated personality scores with wellness measures, diabetes management motivation, and physical activity. Results contributed to an understanding of personality-driven motivations in health behavior change. Summarized findings in a formal report.

Demographics, Data Visualization, Python, Pandas, NLP, Surveys and Reporting

Utah Weight loss Study - University of Utah

2023-2024

This project aimed to find correlations between weight-related metrics and insurance or referral sources. Two clusters of personality types emerged, but the data sample was too small to draw conclusive insights. Results were summarized in a PowerPoint presentation.

Transcription, Data Transformation, Pipeline building, Python, Pandas, NLP, Healthcare, Weight loss

EDII Consultant Finder - BiasProof

2023

This project involved creating a tool that tests employees for various biases and dynamically recommends EDII consultants based on performance metrics. Consultant rankings improve based on client ratings and pre/post-training assessments, allowing the recommendations to evolve and become more effective over time. Summarized results and case studies in a formal report.

ML Model Design and Implementation, Applied AI, Pipeline building, Python, Pandas, NLP, Healthcare