Record friction and failures.
Record what happened and where the decision went off course.
06 / Thinking
Turn experienceinto language for betterdecisions.
I do not leave workplace friction, discrepancies in the numbers, AI failures, or UX uncertainty as passing impressions. I preserve them as decision criteria and bring them back into the next design and implementation.
Read six selected essays ↓01 / OBSERVATIONWhat happened02 / HYPOTHESISWhy it happened03 / CRITERIAWhat to check next04 / NEXT ACTIONReturn it to the design
01 / Numbers
Numbers show where to look next.
STRUCTURE / GAP
01What It Means to Read the Numbers
Look beyond the numbers themselves. Use where they align and diverge to narrow down the underlying structure.
↗KPI / DECISION
02KPI Design, Signal Detection, and Progress Tracking as Core PM Work
KPIs are not just for reporting; they help decide what to inspect and what to change next.
↗02 / Experience
UX designs the sequence of uncertainty.
RESEARCH / QUALITY
03UX Research Quality Starts with Participant Design, Before AI Speeds Up the Work
AI may speed up research, but results are no more reliable when the participant criteria are wrong.
↗INTERACTION / RESPONSE
04From Reading the Web to Interacting with It
Motion is not decoration; it is how an interface responds meaningfully to interaction.
↗03 / AI & Work
AI brings human judgment to the forefront.
FIELD / TOOL
05Why Non-Engineers Should Build Field Tools with AI
People who know the work can encode decision criteria in small tools.
↗PARALLEL / REVIEW
06When Multiple AIs Work in Parallel, Human Review Becomes the Bottleneck
As AI output scales up, the design challenge shifts from generation speed to human review capacity.
↗04 / Judgment assets
Turn articles back into the next piece of work.
Do more than save observations: turn them into constraints, decision criteria, AI instructions, and tool specifications.
Turn them into reusable criteria.
Translate specific examples into conditions and prohibitions that can guide future projects.
Bring them back into design and implementation.
Use them in AI instructions, validation checklists, and tool specifications.
05 / Reading routes
Choose a reading path to fit your time.
Choose what to read based on what you want to learn, rather than browsing a list of article counts.
5 min / Start with the perspective
Read “What It Means to Read the Numbers” and “From Reading the Web to Interacting with It” for a view of how I approach metrics and experience.
15 min / See the way I work
Read the pieces on KPIs, UX research, and field tools to see how I break problems down and connect them to decisions.
30 min / Explore AI operations
Connect parallel AI and human review, the AI operations center, and the decision system for a broader view of personal AI operations.
Ideas are
Preserved to build what comes next.
This archive is not a collection of final answers. It holds hypotheses, decision criteria, and records of failure that can be reused in future projects.