Wyatt Haggard
I study accounting and business information systems at TCU, on the financial technologies track. The coursework is where I learned to read a system; the projects here are where I found out what I actually understood.
Most of them started as a question I couldn't answer by reading about it — how a smart contract behaves once real constraints show up, what it takes to get a model actually serving predictions, how to make a pile of infrastructure data legible at a glance.
None of these are finished products and I'm not claiming expertise in any of them. They're the projects I learned the most from, written up honestly about what worked and what didn't.
Education
Texas Christian University
B.B.A. — Accounting and Business Information Systems: Financial Technologies
Experience
Tax Intern
Prepared individual returns with K-1, rental, and oil & gas income in UltraTax CS; managed depreciation schedules and resolved e-file diagnostics through peak season.
Rodeo Maintenance Worker
Ran tractors, skid steers, and water trucks to prepare arenas and grounds for Pro Rodeo events, and kept irrigation and equipment in working order.
Farmhand
Plowed and tilled 365 acres for seasonal planting, maintained 15 acres of pasture and grounds, and handled routine repairs on farm equipment.
Leadership
President
Elected to lead a 190+ member organization — executive board operations, strategic planning, alumni mixers, and recruitment.
Interests
Tools
Selected work
A blockchain ticketing concept — exploring what it takes to make a ticket a digital asset that can be verified and transferred without a middleman. Mostly a vehicle for learning how smart contracts behave once real constraints show up.
A dashboard for tracking AI infrastructure — capacity, spend, and utilization in one view. Built end to end on my own to see how far I could get turning a messy question into something legible at a glance.
A trained signal model built with course resources, deployed to AWS endpoints and wired to a small hosted app. The interesting part was less the model than everything around it — getting inference to actually serve.
