Clinical AI at enterprise scale
Director of Data Science Engineering at HCA Healthcare, leading engineering, data science, and analytics across critical care, care team optimization, operations, revenue cycle, supply chain, and orthopedics.
Nashville, TN | Applied AI & Machine Learning
Kurtis Pritchard
Data and AI executive with a decade building production machine learning systems in enterprise healthcare, from hands-on architecture to VP-level team and platform strategy.
Current Work
The throughline is pragmatic AI delivery: clinical partnership, measurable outcomes, real MLOps, and organizations that can keep improving after the first model ships.
Director of Data Science Engineering at HCA Healthcare, leading engineering, data science, and analytics across critical care, care team optimization, operations, revenue cycle, supply chain, and orthopedics.
Leading next-generation early-deterioration detection for critical care, with four classification model families designed for risk-stratified alerts more than six hours ahead of deterioration events.
Built and scaled the orthopedic clinical analytics platform from a single-hospital research system into an enterprise platform supporting evidence-based medicine and published research.
Owns platform patterns for real-time clinical feature engineering, model training, registry, evaluation, retrospective analytics, production serving, and model monitoring.
Measured Outcomes
iCORE began as a hands-on build inside an orthopedic research department and grew into a platform supporting clinical meetings, published research, and enterprise analytics.
iCORE-supported evidence-based protocol improvement across the orthopedic program.
Operational and clinical improvement tied to analytics-driven care path refinement.
More patients walking over 50 feet on the day of surgery across the program.
Systems Depth
The work spans model science, production infrastructure, clinical governance, executive communication, and the recruiting loops that make applied AI sustainable.
Classification, forecasting, NLP/LLM, computer vision, risk stratification, evaluation frameworks, and LLM-as-judge patterns.
MLflow, feature stores, model registries, streaming and batch inference, Databricks, Spark, Python, PyTorch, XGBoost, and SQL.
HIPAA and PHI handling, explainability, audit logging, lineage, model risk management, responsible AI frameworks, and clinical advisory review.
Recruiting, technical mentorship, executive communication, product leadership, clinical partnerships, and team formation from zero.
Career Arc
Early iOS products, startup mobile work, clinical software engineering, data science, analytics leadership, and enterprise AI product ownership all show up in the current operating model.
HCA Healthcare
Leads AI-driven product teams across multiple healthcare domains and operates two flagship clinical AI products on shared MLOps infrastructure.
HCA Healthcare
Built a high-performing team of engineers and analysts, expanded beyond orthopedics into neuro and spine, and supported clinical data registry initiatives.
HCA Healthcare
Scaled iCORE across the enterprise and founded the Orthopedic Surgeon Leaders Group with top-performing surgeons.
Southern Joint Replacement Institute
Built the first iCORE application, predictive modeling, analytics platform, dashboards, ETL, infrastructure, and clinical meeting process end to end.
Technical Blog
A writing home for production lessons from patient-safety models, analytics platforms, and AI products that need governance, adoption, and measurable value.
Why evaluation loops, governance, human review, and product accountability matter just as much for agentic systems as they do for regulated AI.
Read essayA practical view of streaming features, retrospective evaluation, clinical event definitions, deployment gates, and operational monitoring.
Read essayHow model validation, advisory cadence, explainability, and workflow design decide whether a patient-safety model becomes real infrastructure.
Read essayPortfolio & Social
Professional background, product work, code, and direct contact channels for applied AI leadership conversations.
AI software portfolio and product work, including the broader CodeSplash ecosystem for developer and builder workflows.