Brenden Runion
Full Stack Engineer & Data Scientist
Denver, Colorado
I build the pipelines that move healthcare data, the models that make sense of it, and the applications people actually use to act on it.
Background
I’m a full stack engineer and data scientist in Denver. For the past five years I’ve worked at Gannett Peak Technical Services on Wyoming’s statewide behavioral health management system - the platform the state runs its Medicaid behavioral health programs on. I build the C# services and SQL Server schemas underneath it and the Python analyses on top of it, the ones program managers use to compare providers and justify multi-year funding decisions.
That range is most of what I like about the job. The same problem is usually a schema question, a data quality question, and a communication question at once, and it isn’t finished until all three are. A claims ingestion pipeline is only useful if the numbers coming out of it survive an audit, and only valuable if someone who has never opened a SQL client can read the report and act on it.
I finished an M.S. in Data Science at the University of Denver in 2025 with a 4.0, on top of dual bachelor’s degrees in Statistics and Sociology from the University of Wyoming. The sociology half gets underrated. It’s where I learned to ask who produced a dataset and what got left out of it before modeling anything - a habit that has caught more real problems in healthcare data than any modeling technique has.
Outside of work I build things end to end and write up what failed alongside what worked. The clearest example is my semantic search engine over 4.4 million Wikipedia articles: the headline finding is that a carefully measured fine-tune lost to the off-the-shelf baseline, and the genuinely useful work was instrumenting why.
What I Work On
Machine Learning & Data Science
Supervised models, NLP and embedding-based retrieval, and evaluation I actually trust - cross-validation, ablations, and error analysis rather than one headline metric.
Data Engineering
HIPAA-compliant ingestion pipelines, schema design that holds up across multiple programs, and query tuning on datasets large enough that the indexing strategy decides whether a report finishes at all.
Full Stack Engineering
Web applications wired to those pipelines end to end, from the schema through the API to the interface. This site included - Vue 3 and Vuetify on a Flask backend.
Right Now
- Extending Wyoming’s behavioral health platform at Gannett Peak Technical Services - claims integration, analytics views, and the reporting state program managers depend on.
- Shipping side projects end to end, most recently a semantic retrieval stack built from the raw Wikipedia dump up.
- Learning Spanish, slowly and stubbornly.
Hiring, collaborating, or just want to talk shop?
Get In Touch