I’m Jimmy Murray—a senior data scientist and AI builder who makes complex systems understandable, governable, and occasionally funny.
↘Projects that combine technical depth, product judgment, and an unreasonable desire to make the answer interesting.
A governed coordination agent for small groups: rotating facilitators, tracking responses, drafting passage materials, and keeping humans in control of consequential decisions.
A growing toolkit for analyzing and generating comedy—topic modeling, phonetic wordplay, assumption-breaking, and performance feedback—built around how working comedians actually write.
An interactive way to see a century of hockey history as a changing neighborhood, translating franchise moves, seasons, wins, and playoff runs into a visual system anyone can explore.
A phonetic modeling experiment that represents names in ARPABET and IPA, aligns syllables, averages consonant and vowel features, and finds the nearest pronounceable result.
My best work lives between “technically difficult” and “people actually need to use this.”
I’ve built data products and AI-enabled applications in federal environments, shaped analytics from messy source data, and translated the results for stakeholders who do not want a lecture on the model.
The comedy background is not an extracurricular footnote. It means I pay attention to timing, assumptions, audience, and whether an idea lands—the same things that separate a clever prototype from a useful product.
NLP, anomaly detection, gradient boosting, clustering, feature engineering
Advanced SQL, wrangling, cleansing, analytics, real-time visualization
React, FastAPI, Streamlit, APIs, agent workflows, human-in-the-loop systems
Databricks, Qlik Sense, PostgreSQL, Plotly, D3, cloud environments
Data science, applied AI, NLP, product-minded engineering—or a problem that does not fit neatly in a job description.