AI-assisted engineering
Agent workflows, explicit control boundaries, durable context, human approval, and evaluations that make automation accountable.
Aaron Fine
Systems engineering for software, infrastructure, and AI-assisted work.
I’m a systems engineer with a software engineering background. I build practical, inspectable systems and explore how AI can support serious analysis without hiding assumptions, provenance, or human judgment.
Current focus
Work that stays understandable
My interests converge on systems whose authority, assumptions, interfaces, and failure modes can be examined rather than merely trusted.
Infrastructure and self-hosting
Maintainable services, open systems, local control, sensible security, and architectures that remain operable after the novelty fades.
Reasoning under uncertainty
Competing explanations, scenario maps, explicit probabilities, adversarial review, and honest scorekeeping when predictions fail.