Professional Summary
I am a Staff-level Data & Analytics leader with over a decade of experience building and scaling data organizations from zero through periods of high growth and transition. I combine deep technical expertise in modern data architecture with a proven ability to lead through executive transitions, procurement cycles, and complex cross-functional initiatives.
My technical foundation is built on scalable warehousing (Snowflake, DuckDB), robust transformation pipelines (dbt, Modal, Airflow), and advanced predictive modeling. I have architected serverless pipelines that parallelize feature engineering across distributed containers, refactored core reporting models to reduce compute costs, and spearheaded enterprise BI rollouts across dozens of global teams.
Currently an MSCS candidate (expected May '26), I split my time between fractional technical leadership, bootstrapping capital-efficient SaaS products, and preparing for a transition into academia to teach the next generation of technical practitioners.
Ventures & Consulting
PDT.dev
Fractional Leadership & ConsultingContext Engineering & AI Integration
A context engineering firm that helps organizations integrate AI into real workflows. We deliver durable process, data, and governance layers in structured 2-week sprints, replacing open-ended consulting with rapid, compliance-ready implementations. You can't layer AI on top of bad data. We fix the data foundation first.
FilingRisk.com
Capital Markets AI
An EDGAR restatement risk prediction system. Utilizes public SEC filings and XBRL tagging, processed through custom LLMs to predict corporate financial reporting risks.
StockStreaks
Algorithmic Trading
An algorithmic trading platform focused on signal generation, quantitative analysis, and automated market strategies.
New Micro-SaaS API wrappers currently in active development.
Teaching Philosophy
Students learn technical material best when they understand not only how something works, but why it matters and what decisions it should enable. In a rapidly evolving landscape, memorizing syntax is a fragile strategy.
I focus my teaching on building practitioner instinct: the ability to look at an unclear business problem, frame the right questions, interrogate imperfect data, and evaluate probabilistic AI outputs responsibly. My instruction is applied, transparent, and iterative. We start with the business problem and the end-user, working backward into the technical mechanics.
Technical skills will inevitably age, but the judgment to know when to trust a system, when to push back on it, and how to use it in service of real decisions will compound over a career.
Publications
The Full Stack Data Playbook
The Full Stack Data Playbook
A Strategic Framework for Fractional Data Leadership
A comprehensive guide moving data teams beyond the "Support Desk" mentality. Drawing on my foundational "How to do Data" framework, this book maps the journey of turning data into a core product feature through four functional phases.
Identifying data sludge and resolving metric drift.
Building lean stacks with DuckDB, Neon, and Turso.
Writing audit-grade SQL and establishing governance.
Operational analytics and resolving LLM context bloat.
Featured Engineering
mini-bedap
HTML / DuckDBThe Context: A prototype Medicare.gov beneficiary dashboard designed to demonstrate full-stack analytics engineering and automated CI/CD deployment.
The Execution: Powered entirely by DuckDB, dbt, and React. It proves out a serverless, in-process OLAP architecture that radically reduces infrastructure bloat while maintaining strict data governance.
Why it Matters: It models how organizations can stay within sustainable computing bounds while scaling to millions of records, turning the "Acceptance Lag" into a signal of system quality rather than a bottleneck.
geminy-cricket
RubyThe Context: As LLMs transition from text generators to autonomous agents executing logic, the risk of unconstrained actions grows exponentially.
The Execution: The Agent Supervision Layer 🦗. An architectural component built in Ruby designed to explicitly monitor, supervise, and control autonomous AI agents executing tasks in production.
Why it Matters: It forces consensus and "Accountant's Rigor" onto agentic AI. You can't let AI manipulate databases or trigger real-world actions without a deterministic safety net. This repo provides that foundational governance.
Availability & Collaboration
I am currently open to conversations regarding Fractional Data Leadership, Context Engineering Consultations (via PDT.dev), and am actively exploring Adjunct Teaching opportunities for the Fall 2026 semester.