Curriculum Vitae

Professional Summary

Data Scientist with 14+ years building and owning predictive models and advanced analytics in SQL, R, and Python across regulated data environments in higher education and national security, plus commercial product development. Own and extend the enrollment prediction framework behind the highest-rated product at a higher-ed analytics firm, reaching 1% MAPE and 59 NPS across 80 users. Delivered two LLM-powered products from prototype to general availability in six months using RAG and custom Python. Co-created the data visualization standards covering research figures and business dashboards, adopted by 28% of a 14,000-person workforce.

Technical skills

Programming: R, Python, SQL, AI-Assisted Development (Claude Code), MATLAB, SAS

LLM: RAG, Pinecone, OpenAI API & Vector Store, Embeddings, Prompt Engineering, LLM-as-Judge Calibration, Golden Datasets

Machine Learning: Scikit-Learn, Predictive Modeling, Model Validation, Feature Engineering, Applied Statistics, Experimental Design, A/B Testing

Visualization: R Shiny, ggplot2, Tableau

Data & Infrastructure: AWS, Docker, Git, CI/CD Pipelines, R Package Development, Data Virtualization (Denodo, Tibco), Spark

Professional Experience

Data Scientist, Capture Higher Ed, Louisville, KY, 08/2022–Present
Designed, implemented, and maintained internal- and partner-facing predictive models in R, Python, and SQL on AWS for a higher education enrollment management and marketing firm. Adapted predictions to institutions of differing size and geography through feature engineering and external data sources including census and school performance data. Enabled Customer Success and education advisors to apply model output to client enrollment strategy, and partnered with 9-12 client institutions to define how each counts enrollment, reconciling those definitions to a common internal standard so predictions stayed comparable across schools.

  • Assumed sole ownership of the enrollment prediction framework and codebase after its original author departed, extending it to become the company’s highest-rated product at 59 NPS across 80 users, with MAPE narrowing from 10% six months out to 1% by May.
  • Scoped 2 LLM-powered products with senior leadership, Product, Engineering, and Infrastructure, translating product requirements into technical approach and delivering both to general availability within a 6-month deadline, with an evaluation framework combining human graders, LLM-as-judge calibrated against human scores, and golden-dataset automated scoring.
  • Found that students visiting campus before applying enroll at near-certainty, redirecting client recruitment spend away from students already converting toward segments where outreach changes the outcome.
  • Engineered R Shiny application automating model validation across the data science team’s full book of 15-20 higher education clients, defining absolute, percent, and raw error measures across 15 prediction points and 5 models, replacing 3 manual checkpoints and surfacing seasonal patterns through exploratory analysis that improved subsequent predictions.
  • Consolidated scattered R scripts into main package infrastructure as a reusable analytical system, eliminating a single-person bottleneck so all 4 team members could generate client onboarding reports and reducing generation time from 2 days to 4 hours.
  • Recognized company-wide by peer vote for first-year impact, selected from 21 candidates across two hiring years (Rookie of the Year, 2023).
  • Assumed ownership of containerized development environment supporting the 4-person data science team’s production model deployment, implementing ARM architecture support to enable migration to faster machines and maintaining annual security updates for R/Python libraries and federal data sources.

Data Scientist, Sandia National Laboratories, Albuquerque, NM, 07/2018–08/2022
Built and deployed interactive applications for 2 predictive models and 4 data visualizations supporting safety, security, and employee retention at a federally funded national security laboratory. Delivered analysis to leadership on the performance of the $177.7 million Laboratory Directed Research & Development (LDRD) program. Streamlined access and data quality for 83 enterprise data sets using SQL and data virtualization (Denodo, Tibco) to join sources across systems into unified data streams, and established research software practices by documenting CI/CD pipelines in Docker and GitLab. Presented on visualization tooling and safety model deployment at the National Laboratories IT Summit (2019).

  • Won competitive CIO funding (1 of 3 from ~20 proposals) to co-develop the enterprise data visualization style guide for research figures and business dashboards, achieving 28% adoption across a 14,000-person workforce as measured by website analytics.
  • Rebuilt the interface for ASSURE for Safety, the laboratory’s safety risk model, partnering with a separate data sciences organization and a UX researcher to replace a legacy .NET implementation with 4,714 lines of R in a deployed R Shiny application that showed managers which organizations faced elevated incident risk and which indicators drove it, validated through two rounds of usability testing.
  • Co-led the 5-person data visualization sub-team supporting a department portfolio of 50+ Tableau and 20+ Shiny visualizations; served as technical lead for the separate 4-person UI/UX data visualization style guide team.
  • Authored 3 internally used R packages as reusable analytical systems, standardizing database connections, plotting defaults, and package management.
  • Evaluated Tableau against R Shiny for enterprise visualization delivery, comparing cost, licensing, version control, and customization, and presented the findings to IT staff across the national laboratory system.

Data Analyst, HID Global, Albuquerque, NM, 01/2012–06/2018
As a member of HID Global, a leading provider of secure identity solutions specializing in products and services that enable trusted access to physical and digital spaces. I managed the human testing lab; coordinated 53 research studies, in various sites in the US to evaluate biometric sensors. Analyzed resulting data to measure biometric sensor performance and prepare it for machine learning algorithms. Oversaw 12.89 terabytes of collected data. Presented research findings using various media (print, video, and webinars) to both technical and non-technical staff, and clients. Researched and identified new methods for spoofing biometric devices.

  • Diagnosed 3 unknown security threats to HID Global biometric sensors that drove improvements to the fake finger detection algorithm; collaborated with hardware engineers and computer scientists on R&D for 4 sensors, including one for guests at the most visited vacation resort in the world.
  • Programmed a data pipeline processing 3,000 wireless security camera videos collected daily across US research sites for an object detection algorithm.
  • Cataloged collected data in a MATLAB accessible MySQL database and devised new filename conventions, expediting algorithm development and improving data quality across research sites.
  • Cut research study cost and product development time by improving subject recruitment, instituting new recruitment methods and strengthening communication to reach a 48% subject retention rate.
  • Oversaw 12.89 terabytes of collected data across US research sites.

Data Analyst, NMSU Department of Public Health Sciences, Las Cruces, NM, 08/2010–05/2011
Provided statistical consulting to 8 graduate students and 11 faculty members across the full research lifecycle while attending classes full-time: grant writing, hypothesis development, survey and experimental design, data preparation, analysis in SAS and R, and interpretation of results for presentation and publication.

  • Co-authored a presentation with the university’s associate dean of research applying factor analysis to construct a new metric for measuring hooking-up behavior among college students.
  • Position created by the department head following demonstrated work in applied statistics, SAS and R programming, and communication with non-technical researchers.

Student Statistical Consultant, NMSU Department of Economics, Applied Statistics, and International Business, Las Cruces, NM, 08/2010–05/2011
Advised graduate students and university staff on statistical methods and software selection, analyzing data in SAS and R across projects on MBA admissions criteria, motel crime rates, and online dating behavior.

Intern Statistician, USDA Food Safety Inspection Service, Washington, DC, 04/2010–08/2010
Ensured data quality for raw data collected across US meat, poultry, and processed egg facilities, cleaning source data and updating the data warehouse. Applied statistical methods to estimate foodborne pathogen prevalence and prepared figures for professional presentations.

  • Authored a public-facing report on pathogen prevalence in raw poultry products.
  • Consulted with interns from non-technical backgrounds on appropriate statistical technique for their projects.

Education

MS in Applied Statistics, New Mexico State University, Las Cruces, NM, 2011

Coursework in experimental design, regression and linear models, multivariate analysis, sampling methodology, and statistical consulting; programming in R and SAS.

BS in Biology (Minor: Applied Math), New Mexico State University, Las Cruces, NM, 2009

Certification

Fundamentals of Computing Specialization (7 courses), Rice University via Coursera. 07/2018 (Verify)

Applied Data Science with Python Specialization (5 courses), University of Michigan via Coursera. 09/2017 (Verify)

Professional Training

Cloudera Data Scientist Training by Cloudera. 02/14/2022–02/17/2022

SAFe® Product Owner/Product Manager by Scaled Agile. 01/11/2022–01/13/2022

Storytelling with Data by Cole Nussbaumer Knaflic. 10/15/2020

JavaScript for Shiny Users at RStudio Conference 2020. 01/27/2020–01/28/2020

Apache Spark™ Programming (DB 105) by DataBricks. 09/24/2019

Presentations

Viz Wars: Tableau vs. Shiny, with Patrick Carlson. National Laboratories Information Technology Summit, Boise, ID. 05/2019

Visualizing a Safety Model with R Shiny, with Scarlett Marklin. National Laboratories Information Technology Summit, Boise, ID. 05/2019

Conferences

Posit Conference 2025 by Posit, Virtual. 09/16/2025.

Posit Conference 2024 by Posit, Virtual. 08/12/2024.

Posit Conference 2023 by Posit, Virtual. 09/19/2023.

RStudio Global Conference 2022 by RStudio, Virtual. 07/25/2022.

RStudio Global Conference 2021 by RStudio, Virtual. 01/21/2021.

DOE Virtual Data Days 2020 by The Department of Energy, Virtual. 10/05/2020–10/07/2020.

Spark + AI 2020 Summit by DataBricks, Virtual. 6/22/2020

Conference on Data Analysis by Los Alamos National Laboratory, Santa Fe, NM. 02/25/2020–02/27/2020.

RStudio Conference 2020 by RStudio, San Francisco, CA. 01/29/2020–01/30/2020

National Laboratories Information Technology Summit by Federal Business Council, Inc. Boise, Idaho. 04/28/2019–04/31/2019