Reliable AI and agents
How evidence, boundaries, and observable execution turn model capability into dependable systems.
Writer · researcher · systems practitioner
Making complex systems legible.
I write short essays about the boundaries, evidence, and operating habits that make technology—and the work around it—more reliable.
Read the writing →Recurring questions
Themes in the work
Infrastructure and recovery
Why backups, restore drills, deployment contracts, and explicit ownership are part of product design.
Knowledge and education
How tools can support learning, synthesis, and durable intellectual work without hiding the process.
Language and rhetoric
How names, metaphors, and explanatory structure shape what a technical system permits us to see.
Professional work
Systems that can explain themselves
My work centers on designing, evaluating, and operating systems whose decisions and failure modes remain inspectable.
View the professional CV →