I work on decision systems where ambiguity has consequences.
My focus is not model performance or feature optimization. It is how authority, accountability, and meaning behave once automation enters environments governed by oversight, escalation, and post-incident review.
Most failures I see are not technical.
They are architectural.
They occur when:
- decision authority becomes diffuse
- responsibility is implied instead of assigned
- human involvement exists in form, not function
- systems optimize locally while intent erodes globally
I study how these failures emerge—and how to design systems that resist them.
My background spans AI and machine learning, systems analysis, and technical and strategic writing in regulated, high-accountability contexts. I operate across engineering, governance, and decision-support layers, focusing on how choices are framed, constrained, reviewed, and defended.
I don’t build optimism into systems.
I build explainability, stoppability, and ownership.
If a decision cannot be clearly traced, justified, and interrupted, it does not meet the standard—regardless of performance metrics.
I work selectively, on problems where decisions must hold up not just at deployment, but afterward.
Orientation
My work focuses on how authority, meaning, and control degrade inside high-tempo, AI-enabled decision systems - often without visible failure. I write about interpretive drift, cognitive bandwidth collapse, ISR fusion, and post-hoc control illusions because these are not theoretical risks.
They are structural failure modes that emerge when systems behave correctly, oversight appears intact, and governance arrives too late to matter. I approach these problems at the architecture layer. Not interface design. Not model internals.
Architecture as the boundary where authority is encoded, constraints are enforced - or quietly omitted - and human cognition is forced to operate under speed.
To sharpen that layer, I’ve enrolled in Carnegie Mellon University’s certificate program focused on systems architecture, operational risk, and resilience in complex socio-technical environments.
CMU’s emphasis on systems architecture and resilience deepens my ability to identify where constraints must be embedded structurally - not just described retrospectively.
CMU’s work treats failure as emergent, risk as structural, and humans as part of the system - not an external check.
This is not a pivot or a rebrand. It is reinforcement. My writing remains oriented toward decision integrity, constraint design, and the conditions under which control either exists in time to matter - or becomes a narrative applied afterward.
If explanation feels reassuring but nothing can be stopped, the system is already telling you where control actually lives.