About :: Adrian Vance Writes

About

Adrian Vance writes about applied AI, model evaluation, and decision systems for technical audiences.

His background is operations research: roughly a decade of building optimization models, allocation frameworks, and decision support tooling for large organizations, in Python, R, and SQL. That work produced a specific interest, which is the gap between a model that is technically correct and a system that a decision maker will actually use.

He writes about that gap. Evaluation harnesses, objective functions, retrieval design, and the organizational habits that make analytics credible or discardable.

He works with technical content teams at data, analytics, and AI infrastructure companies. He does not write listicles.

Topics covered

LLM evaluation, eval harness design, LLM-as-judge validation
Retrieval, context assembly, and memory architecture for AI applications
Optimization and multi-objective decision modeling
Analytics credibility: why models get rejected by stakeholders
Python, R, and SQL tooling for analysts