Hi, I'm Cole.
These days, my focus is teaching machines to read cuneiform, the world's oldest writing.
Previously, I spent 6 years as Co-Founder & CTO at AbstractCRE, an enterprise product for extracting structured data from commercial real estate legal documents.
Why cuneiform
- There's a lot of it.400,000+ cuneiform texts bear witness to 3500 years of history, from the invention of the wheel (and writing itself) to the Roman Empire. These texts – administrative texts, personal letters, royal inscriptions, and more – give a view from the ground during some of the most foundational events in our past.
- We have a lot left to learn.With so many texts and so few specialists with the skills to read them (Assyriologists), progress has been slow in understanding what these texts have to tell us about the story of human civilization. Basic digital infrastructure and new computational methods promise a new way of working which will dramatically accelerate this project.
- These texts are largely inaccessible to the general public.Because an exceedingly small proportion of the corpus has been translated, years of specialized training is effectively a prerequisite for engaging with the first half of human history. (And some Assyriologists are baffled why there isn't a greater cultural awareness...) My goal is to make this rich material accessible to everyone. I'm dedicating years to learning these languages so no one else has to.
- I believe I have a moral duty to work on these problems.I believe the skills I built at AbstractCRE uniquely position me to work on these problems. Turns out, whether you're working with legal documents 4 years old or 4000 years old, the solutions needed aren't all too different.
What I'm up to
DPhil Asian & Middle Eastern Studies
University of Oxford · Starting October 2026
Building NLP infrastructure for cuneiform. The document-understanding problems are surprisingly similar to the ones I spent six years solving — just a few millennia older.
MPhil Cuneiform Studies
University of Oxford
Expected completion: June 2026. Two years of intensive Akkadian and Sumerian. You can't build good computational tools for a domain you don't deeply understand, so I'm learning to read these texts the old-fashioned way first.
B.S. (Honors) Symbolic Systems
Stanford University
Concentration in AI. Minor in Classics. Came back six years later to finish what I started. SymSys — the study of how humans and machines represent and process information — turned out to be the perfect degree for what I'm doing now.
Independent Contractor
Lush Software LLC
A deliberate pause after six years of startup intensity. Took contract work while figuring out how to spend the next decade.
Co-Founder & CTO
Enterprise SaaS product to extract structured data from messy commercial real estate documents, run reports, and connect it to other systems of record. Architected, built, and maintained the entire stack for six years: user-facing product, internal annotation platform, data pipelines, custom ML models.
Undergrad (Pt. 1)
Stanford University
Leave of absence to start AbstractCRE
Born and Raised
North Georgia mountains
Grew up in moonshine country, taught myself to code at 14, and talked my way into working for startups while I finished high school online.