Director of Product & Technology

Edudyn, Inc. (EDI) · joined as Product Development Specialist · more

• Sole developer and maintainer of a new platform for an EDI client that aligns provider time cards and services with student enrollment records, replacing an earlier AI tool for parsing unstructured notes and dramatically reducing billing turnaround.
• Built two new StudentTrac products: an AI-powered external transcript import that parses, maps and imports prior coursework with staff review in a side-by-side view, and an SMS communication platform for staff, guardians and students designed for automated nudges.
• Oversee StudentTrac maintenance and perfective work: fixes for user-reported issues and small feature requests.
• Built a family of cloud agents in Google Chat and Jira that investigate help desk tickets, build and test fixes, classify tickets and create them from customer chats; set best practices and a starter toolkit for agent use on the codebase.

Data Science Consultant

Beam Data

• LinkedIn engagement automation. Tech lead for prompt engineering, agent development and the front end of a multi-agent system that finds an audience, researches and writes content, edits it for compliance, and manages engagement on a schedule.
• Business development automation. Sole responsibility for the multi-LLM agent structure and front end of a tool that rates the fit of potential clients and writes pitches, built on refined web-scraping pipelines.
• Identity verification for a mobility startup. OCR and LLM tools that extract information from legal documents from any North American jurisdiction and score the quality of what was extracted.

LangChain · LangGraph · CrewAI · Streamlit · scikit-learn · AWS Bedrock, Textract, Rekognition · MongoDB · open-weight LLMs

Research Analyst, Prompt Engineering

Welocalize

• LLM research-agent evaluation. Wrote the rubrics and standards for prompts and for evaluating research-agent performance, handed to other teams to build high-quality datasets.
• Causal reasoning. Wrote prompts, responses and guidelines that probe cause-and-effect relationships in narratives, to advance causal reasoning in large language models.

Associate Professor of Philosophy, with tenure

University of Oklahoma · see the academic CV

The Reasoning Project

A laboratory for doxastic norms

Language models do small, isolated jobs and deterministic code decides what stands, so I can test how well models carry epistemic norms. Pre-registered experiments, about 3,000 model-free tests.

Calling the peloton

Calibrated predictions for pro road races

A LightGBM model stack and a nightly crew of agents that publish predictions for WorldTour and ProSeries races, then settle every call in public.

Homegreens

A standardized index of green coffee

Creator, with full-stack responsibility. LLM extraction and cleaning across 20 sellers and 2,762 coffees, with an evaluation set for the flavor classifier.

Mapping Hume’s Treatise

Quotation profiles for a historical text

NLP over 101 scholarly papers from PhilPapers’ index of work on Hume and personal identity, profiling how each one cites the Treatise, with a public web front end for students, teachers and scholars.

Associate Professor, with tenure

University of Oklahoma

Assistant Professor

University of Oklahoma

Visiting Assistant Professor

Virginia Tech

Ph.D. in Philosophy

University of North Carolina at Chapel Hill

Committee: Alan Nelson (chair), Robert M. Adams, Simon Blackburn, Ram Neta, Geoffrey Sayre-McCord.

M.A. in Philosophy

University of Wisconsin–Milwaukee

B.A. in Philosophy, Politics, and Economics

Claremont McKenna College

Locke’s Science of Knowledge

Routledge · monograph