Summary
Breck Baldwin is a founder, applied LLM/AI researcher, and system architect with over a decade of experience building reliable, production-grade AI across defense, medicine, finance, and entertainment. He specializes in making large language models predictable and testable—publishing on LLM determinism and inventing "Cyber Linguistics" to treat model I/O as structured interfaces rather than black boxes. A former PI on DARPA and NIH projects who built LingPipe (3,000+ citations) and led AI initiatives at Comcast, he combines deep academic credentials (PhD, prolific publications, two books) with hands-on product delivery. Breck runs a stealth startup focused on agentic evaluation and LLM behavior, and is known for pragmatic systems: modular, low-code, human-in-the-loop architectures that scale in messy real-world settings. An engaging collaborator and public-facing communicator (host of a local AI radio show), he brings both technical rigor and a knack for leaving organizations better than he found them.
9 years of coding experience
24 years of employment as a software developer
Bachelor's degree, Applied Linguistics, Bachelor's degree, Applied Linguistics at Hampshire College
PhD, Computer Science, PhD, Computer Science at University of Pennsylvania