R. Jerome Dixon

Professional dossier

R. Jerome Dixon

Applied Data Scientist · MLOps Architect · PhD Researcher

Innovative applied data scientist and MLOps architect with more than 20 years of experience in analytics and operations research. Retired U.S. Navy logistician with expertise in healthcare and supply-chain analytics. Skilled in designing serverless architectures, survival and risk models, and ML/AI solutions on AWS.

jerome@mushinsolutions.com · ORCID 0000-0001-8622-0597

Skills

  • Languages: Python, R, TypeScript, JavaScript, C++/CMake, Linux Bash.
  • Cloud (AWS): Lambda, S3, Fargate, EC2, API Gateway, Cognito, DynamoDB, Bedrock, EMR, Athena, Glue, QuickSight, Step Functions, CloudWatch, Route 53.
  • Machine learning & AI: XGBoost, CatBoost, Generative AI / LLMs (Ollama, Phi-3 Mini, AWS Bedrock), AWS Comprehend, PyTextRank, spaCy, SHAP, Formal Feature Attribution (FFA), BupaR process mining, Dynamic Time Warping (DTW), FP-Growth.
  • Data architecture: DuckDB, Parquet, serverless graph data lakehouse.
  • Certifications: AWS Certified Solutions Architect – Professional, AWS Certified Data Analytics, AWS Certified Machine Learning.

Honors

  • Virginia Beach Bio Innovation Challenge Winner, Center for Advancing Innovation (2019)
  • U.S. Census Bureau Phase I Supply Chain Challenge Winner (2023)
  • Foreign Area Officer Association Award for International Affairs, Joint Forces Staff College (2012) — winning paper: Afghanistan – A New Economic Model for Change

Professional experience

Senior Operations Research Analyst

CANA LLC · Mar 2017 – Present

  • Engineered the Supply Chain Open Source Intelligence (SC-OSI) serverless graph data lakehouse for Earth 616, a Department of the Air Force / AFRL strategic program for supply chain analytics, digital asset traceability, and blockchain, using S3, Parquet, DuckDB, Lambda, and AWS Fargate.
  • Integrated generative AI to automate supply-chain risk reporting (AWS Bedrock / Amazon Nova Micro and on-device Ollama Phi-3 Mini).
  • Built multi-source NLP pipelines (AWS Comprehend, PyTextRank) from DLA FLIS, FAA PMA, DLA SA-SOR, and Gmail alerts into a supply-chain risk ontology dashboard.
  • Built persona-driven React and serverless dashboards for SCRM analysts, PEOs, and geographic operational commanders.
  • Migrated a USMC Power BI application to AWS QuickSight with Lambda and Step Functions.
  • Won the U.S. Census Bureau Phase I Supply Chain Challenge; competed in the Palantir/NIH Long COVID Computational Challenge.

Data Scientist / Backend Data Engineer

SurgicalED VR · Sep 2023 – Present

  • Led an AWS account migration and GitHub Actions OIDC CI/CD for a multi-tenant React / TypeScript / DynamoDB / Cognito SaaS portal.
  • Architected serverless pipelines that turn Unity VR assessment JSON into partitioned Parquet via Lambda, Glue, and Athena.
  • Applied single-subject design and learning-curve theory with DTW, log dimensionless jerk, and OSAT composite skill assessments.
  • Authored analysis workflows and technical documentation for multiple VR surgical-training platforms.
  • Engineered HTML 3D trajectory visuals generated via Lambda and embedded in QuickSight dashboards.

PhD Researcher

VCU Integrative Life Sciences · Mar 2018 – Present

  • Designed an end-to-end ML pipeline to predict opioid-related and polypharmacy high-cost geriatric ED events using the Virginia All-Payer Claims Database (APCD).
  • Developed CPIC pharmacogenomic time-to-event models for drug–drug and gene–drug contributions to fall and ED risk.
  • Trained XGBoost and CatBoost ensembles with SHAP, Formal Feature Attribution (FFA), FP-Growth, and BupaR process mining.
  • Created and deployed a serverless clinical decision-support risk dashboard (Docker, ECR, API Gateway).
  • Used Optuna and Platt calibration on highly imbalanced clinical datasets.
  • Authored five manuscripts at the intersection of explainable AI, precision analytics, and PGx risk prediction (Clinical and Translational Science; CH1 invited at Journal of Personalized Medicine, pending formal acceptance; CH2 journal pending).

Data Scientist (Part-Time)

UVA School of Data Science · Sep 2022 – Mar 2026

  • Developed machine-learning workflows for pediatric heart-transplant graft loss in support of UVA Health Pediatrics.
  • Used gradient-boosted trees, hyperparameter optimization, and SHAP explainability.
  • Presented at the International Society for Heart and Lung Transplantation (ISHLT) Annual Meetings in Denver (2023) and Prague (2024), and co-authored pediatric transplantation studies.
  • Applied CatBoost-Cox, XGBoost-Cox, and Accelerated Oblique Random Survival Forests to graft-loss analysis.

Data Consultant

Virginia DBHDS · Jan 2017 – Jan 2022

  • Data and program-policy consulting for the SAMHSA Youth Substance Abuse and Treatment (YSAT) grant.

Publications & research

Peer-reviewed papers and manuscripts in review. Covers and DOIs match jerome3590.github.io and ORCID 0000-0001-8622-0597.

Pharmacogenomics & claims analytics

Virginia Commonwealth University · with Elvin T. Price

Surgical simulation & virtual reality

VCU · Old Dominion University

Pediatric heart transplantation

University of Virginia School of Data Science

Earlier work

  • Modeling and Simulation of Healthcare Systems · Oct 2019

    Using System Dynamics with Machine Learning for Modeling Patient Flow and Hospital Staffing Requirements

    Chapter 1 in Modeling and Simulation of Healthcare Systems (Selected Papers on System Dynamics Collection). KDP Publishers.

    doi.org/10.13140/RG.2.2.23651.86564/1

  • NAVSUP · Feb 2012

    Supply Chain Event Management (SCEM)

    Dixon RJ. Naval Supply Systems Command Business Systems Center. Pp. 25–27. Government periodical; no DOI.

    ResearchGate 325228789

Education

  • PhD in Integrative Life Sciences (in progress), Virginia Commonwealth University (2018–present)
  • Master of Decision Analytics, Virginia Commonwealth University (2016–2018)
  • M.S. in Information Technology Management, Naval Postgraduate School (2005–2007)
  • B.S. General Engineering, United States Naval Academy (1993–1997)

Early career & military leadership (1997–2016)

  • Director, F/A-18 Life Cycle Logistics, Defense Logistics Agency (2013–2016)
  • IT Project Officer, Navy Supply Business Systems Center (2009–2013)
  • Readiness Officer, USS John C. Stennis (2007–2009): supervised 400+ personnel; $120M budget and $500M in material assets
  • Contracting/Ordering Officer, Coalition Provisional Authority, Basrah, Iraq (2003–2004): 320 projects, over $30M
  • Air Logistics Officer, Naval Central Command / CTF-53, Bahrain (2004–2005): passenger, cargo, ammunition, and mail planning for Fifth Fleet AOR
  • U.S. Navy Supply Corps Officer (1998–2016)
  • United States Naval Academy (1993–1997)