AI Orchestrator & ML Engineer building healthcare AI that bridges the gap between clinical reality and machine intelligence. Anthropic & AMD AI Certified.
I spent 8+ years on the clinical frontlines — as an anesthesia tech at Queen's Medical Center in Honolulu and in inpatient psychiatry — learning to read patients, patterns, and urgency in real time.
That clinical intuition now drives everything I build. I saw firsthand how documentation steals time from patients, how diagnoses get delayed, and how cognitive overload is the default in healthcare. I pivoted into AI/ML to fix that.
Today I build healthcare AI systems — from RAG-powered clinical decision support to hospital readmission prediction pipelines — with a relentless focus on interpretability, fairness, and real-world clinical impact.
Deep clinical background in anesthesia & psychiatry informs every ML system I design.
Building intelligent agents with Claude API, MCP protocols, and multi-agent architectures.
End-to-end pipelines from data engineering to model deployment with SHAP interpretability.
Fairness auditing, bias detection, and transparent model decisions are non-negotiable.
Healthcare-focused AI and ML systems, from clinical NLP to predictive analytics — each solving real problems at the intersection of medicine and technology.
End-to-end ML pipeline for predicting 30-day hospital readmissions. Covers synthetic EHR data generation, preprocessing, model development with hyperparameter tuning, SHAP-based clinical interpretability, and fairness auditing across demographics.
View on GitHub →Healthcare AI Education Agent combining Meta's V-JEPA 2 with Claude AI. Analyzes medical procedure videos to generate step breakdowns, teaching narration, quizzes, and safety notes for clinical education.
View on GitHub →Full RAG pipeline with Mistral-7B for medical knowledge retrieval. Enables clinicians to query medical literature and receive contextually grounded, evidence-based answers for clinical decision support.
View on GitHub →Comprehensive analysis of chronic disease prevalence across the United States — diabetes, obesity, heart disease, and physical inactivity — with interactive visualizations and geospatial mapping.
View on GitHub →Regex-powered extraction of clinical concepts — ICD-10 codes, CPT codes, medications, and vitals — from EHR clinical notes using Databricks SQL for scalable healthcare data processing.
View on GitHub →Predictive maintenance system using neural networks to forecast wind turbine component failures — reducing downtime and optimizing energy production through data-driven maintenance scheduling.
View on GitHub →Certifications and education that back up the work — from cloud fundamentals to cutting-edge AI agent architectures.
A practitioner's toolkit — refined through building real systems, not just tutorials.
A non-linear career fueled by curiosity and a refusal to accept the status quo in healthcare.
Building agentic healthcare AI systems, contributing to open-source, and pursuing the UT Austin AI/ML certificate. Attending AI DEV 26 x SF and AMD AI DevDay in April.
Completed 12 Anthropic Academy courses, earned Microsoft Azure & AI certifications, AMD AI Academy, and Databricks badges. Built end-to-end ML pipelines and RAG systems.
Taught myself Python, then dove into machine learning — applying the same empathy used to understand students and patients to understanding data and models.
Anesthesia Tech at Queen's Medical Center, Honolulu. Inpatient psychiatry. Special education. Learned to read patients, monitors, and patterns under pressure — the foundation for everything I build today.
Open to ML Engineer, Healthcare AI, Clinical AI, and Developer Education roles. Remote or relocating to Seattle, SF, or NYC.