Adam W. Stauffer
Finance and economics lecturer. Former ETF trader and CIO. I build courses, and other things, with AI.
I teach finance and economics at the Shidler College of Business, University of Hawaiʻi at Mānoa, to undergraduate, MBA and executive MBA students in Honolulu, Vietnam and online. Before teaching, I was a trader and market-maker in U.S.-listed ETFs at Barclays Capital and Lehman Brothers, founded and ran Springline Capital as Chief Investment Officer, and launched Advent Software's global solution for hedge funds and fund administrators. I hold an MBA in Finance from Wharton and was a CFA charterholder from 2004 to 2011.
Every course project I teach is built around one problem: AI now drafts the first pass that junior analysts used to build, and that first pass was where they learned judgment.
I taught myself Claude Code, Codex and GitHub, and build everything below with them.
ResumeResume (PDF)CVBioLinkedInGitHub
Writing
The doer–reviewer dilemma
What happens to learning, judgment and career advancement when AI does the first draft, and how I've rebuilt my courses around it.
October 2026.
Teaching reports
Lessons on teaching and grading with AI from four case runs
What four runs of AI-permitted business cases showed about grading, disclosure and where students actually learn, and what I'm changing next.
October 2026.
Books in progress
The 2008 Global Financial Crisis: An Extended Case in Macroeconomics
Told through the crisis chronology and from the ETF desk at Lehman Brothers and then Barclays Capital, where I traded through it.
Lecture decks built.
The 2026 Strait of Hormuz Closure: An Extended Case in Microeconomics
Supply and demand, costs, market structure and externalities, taught through one live shock to the world's oil supply.
Lecture decks built.
AI + GitHub + Humans = Future of Work: A Rethinking of Higher Education Curriculum
Courses for a world where AI drafts the first pass: students frame, specify, build with AI, audit and decide, in the open on GitHub. It starts from the doer–reviewer dilemma.
Cases in progress.
Applied AI work
Kumu: AI-first course tutorial site
Stage-by-stage case projects across four subjects, with checklists, self-quizzes, hands-on labs and an AI tutor. Every project runs Frame → Specify → Build with AI → Validate → Decide, in the open on GitHub.
Course materials
Syllabi, staged project briefs, templates and the GitHub & AI and Claude appendix decks my students use.
Builds
Organic Farm Management & Certification Platform
Software for organic farms built around certification: it turns certifier forms into structured workflows, with AI agents extracting the forms, and walks a farm through its Organic System Plan. Next.js and FastAPI on PostgreSQL, built since December 2024 with Claude Code and Codex.
Private repository.
Treehouse: design and build
A real treehouse design record: AI-drafted studies and drawings, my rulings dated, not for construction. A treehouse in an oak in the Great Lakes region. AI drafts the studies, the 13-sheet construction set and a real-time walkthrough from one geometry model; I rule on every option, and each ruling is numbered and dated.
Greenshoot.org
A disaster-recovery registry I founded and built after Hurricane Irma, connecting survivors' needs with donors. Founder Institute graduate.
Regenerative food forest
A one-acre, multi-story food forest on former pastureland, with a soil-amendment calculator built on Albrecht's and Kittredge's methods.