Systems-level leverage
Reducing coordination cost
Over time, I’ve noticed that many design problems stop being design problems. They become coordination problems between teams, systems, documentation and decisions made months or years apart.
-
Patterns are redesigned instead of reused
-
Decisions are revisited because rationale is lost
-
Inconsistencies accumulate faster than reviews can catch them
-
Human attention is spent on checking, not solving
What leverage looks like in practice
-
Creating shared interaction patterns that outlive individual teams
-
Making system state visible so users don’t have to infer it
-
Shifting effort from repeated review to upfront structure
-
Using automation or AI to surface issues, not decide outcomes
Recently, I’ve been exploring where AI can help at this layer; not to generate interfaces, but to support consistency, documentation and drift detection across systems that are becoming too complex for manual oversight.
I’m most interested in uses where AI reduces coordination cost while preserving human judgment, rather than replacing it.
This work doesn’t replace product or interaction design, it complements it. Especially in organizations where the hardest problems aren’t about what to design but how to keep systems coherent over time.