Port Angeles, Washington, United States
I build practical software and data tools for operational and product questions where correctness, traceability, and a working implementation matter. My work spans applied AI, data-intensive applications, predictive and segmentation workflows, forecasting, and decision support. I’m comfortable moving from an ambiguous question through data design and modeling to a working application, API, or repeatable analytical workflow. I work primarily in Python, SQL, and R, using FastAPI, PostgreSQL/pgvector, React/TypeScript, Docker, and AWS where they fit. Through Praxish, I’m building independent software, including a private-beta career evidence system that turns resumes, notes, and interactions into traceable profiles and tailored documents, and Cinnaname, an interactive name explorer built from U.S. Social Security name histories, embedding-based connotation scoring, and lexical data. Previously, I led Advanced Analytics & Data Science solutioning at Prophet and managed data science and analytics work at Capgemini Invent. That work included segmentation systems, scenario forecasting and Monte Carlo simulation, donor propensity modeling, patient-journey analysis, CRM attribution, and executive decision support. I’m available for contract engagements and selected remote roles. I’m most useful when the problem is real, ownership is clear, and direct communication and written, asynchronous progress are valued. praxish.com
Independent applied AI and data practice, operated through Storm King LLC, focused on end-to-end product development and selected contract work. • Building a private-beta career evidence system that turns resumes, notes, and interactions into versioned profiles, grounded Q&A, role matching, tailored documents, and portable exports. • Designed the system around FastAPI, PostgreSQL/pgvector, SQLAlchemy/Alembic, database-backed asynchronous workers, retries and deduplication, workspace access controls, artifact tracking, and usage/cost telemetry. • Built and deployed Cinnaname, a React/TypeScript and FastAPI/PostgreSQL explorer combining U.S. Social Security name histories, embedding-based connotation scoring, WordNet analysis, and interactive Plotly views. • Available for focused remote contract work involving applied AI, data products, internal tools, automation, predictive analytics, and decision support.
Led Advanced Analytics & Data Science solutioning, technical design, reusable methods, delivery, and team development. • Built a reusable R/Python pre-segmentation pipeline covering redundancy, dimension reduction, separability, and stability checks. On typical builds, it reduced the candidate model space by approximately 99.6%, shortened cycles by weeks, and saved an estimated 100 associate-hours per project. • Productized model delivery through reusable R workflows, Shiny scoring tools, Python inference, and batch and single-respondent utilities. • Developed a donor-propensity model combining institutional, engagement, Experian, and DonorSearch data; optimized for recall using F2 and surfaced previously unsolicited high-propensity prospects for nurture. • Scoped analytics engagements, mentored associates and managers, and translated technical work for client and executive audiences.
Managed data science and analytics work across strategy and operations. • Built 1.5–3-year scenario forecasts and Monte Carlo simulations for a global PC and printing OEM, separating temporary COVID demand displacement from structural shifts and supporting C-suite planning. • Combined claims, pharmacy, and operational signals to map a biopharma patient journey, identify bottlenecks, and support weekly executive briefs and quarterly business reviews. • Developed CRM and contribution-attribution analysis across cloud sales teams to assess how meetings, specialist engagement, and account coverage related to pipeline progression.
Six-month post-acquisition contract focused on analytics priorities and operating risks. • Assessed business-unit priorities and data opportunities for a developing analytics capability. • Built early epidemiological scenarios that flagged COVID-related risks to cell-therapy workflows, inpatient capacity, and hospital operations. • Translated findings into strategy recommendations on supply and observation constraints.
Led consulting data science capabilities and reusable offerings across solution design, sales support, delivery, and coaching. • Scoped and delivered predictive modeling, market insight, and segmentation work. • Developed reusable analytical methods and intellectual property to accelerate client delivery. • Mentored consultants on technical work, project execution, and career development.