№ 01
Structure creates observability
Preserve continuity, attribution, and timing, and the patterns surface on their own.
No. 01 · The Logic Garden
I design systems that fit into how people already work.
Workflow systems, automation, and decision architectures built so people can stop working around their own tools.
Solutions Architecture
Ambiguous requirements shaped into systems people can trust and explain.
Automation & AI
Automation paired with structure, so it gets relied on instead of bypassed.
Data Modeling
Structure that preserves what a record means, not only what it stores.
Design principles
What makes a system worth trusting?
Six ideas underneath every system I build.
№ 02
Human behavior is the environment
It's the medium a system lives inside, not noise to design around.
№ 03
Friction must justify itself
Protect the friction that creates intention. Cut the administrative kind.
№ 04
Meet people where they work
People bring habits and tools they already trust. Build on those instead of fighting them.
№ 05
Trust breaks before the system does
A system loses people the moment it gets hard to explain. That's when shadow workflows appear.
№ 06
AI as an augment
It enriches a system that already works; it doesn't stand in for one.
Featured systems
Field specimens.
Specimen 01Medium
The Hidden Join
Provider and Location Routing Without Relational Support
Synthesized a join key within a platform that didn't support joins to route patient feedback to the correct review page.
Data ModelingConstraint-Based LogicMany-to-Many Relationships
Google Sheets · CSV/JSON Import · Custom Field Mapping
Specimen 02Advanced
The Storyteller Engine
Structured Synthetic CX Data for Testing and Demos
Generates structured, relational datasets that simulate real-world customer behavior for testing and demos.
Relational ModelingSchema DesignConstraint-Based LogicValidation Logic
Google Sheets · Mockaroo · Apps Script
Specimen 03High (Cross-System)
Three Systems, One Contact Record
How Persistent State Broke Attribution
Prevents cross-system data contamination by enforcing interaction-level state instead of shared contact data.
Data ModelingConstraint-Based LogicState Management
Multi-System Integration Environment · APIs · Contact-Based Survey Platform · CRM
Field notes
A living collection.
Observations, experiments, and the decisions behind the systems. Essays that get written when they are ready, not on a schedule.
All field notes →SystemsTwo weeks ago
When the score measures the wrong unit
An average that hides its own spread will agree with the data and mislead the reader at the same time.
6 min read
ProcessLast month
Reading workarounds as requirements
The spreadsheet someone runs on the side is usually the actual spec. Here is how to mine it.
5 min read
Automation & AIEarlier this year
Where a human belongs in the loop
A short test for which decisions to automate and which to hand back to a person.
7 min read
DataThis year
Naming things before modeling them
Six weeks of vocabulary work saved a migration that would have modeled the wrong world.
8 min read
Explorations
What’s taking root.
Sprout № 01
The Living Archive
An ongoing collaborative story, archived so continuity, attribution, and timing stay intact, and the chemistry between its characters surfaces as pattern you can study.
Sprout № 02
The Taste Model
Collaborative listening and taste modeling over a 2,461-album K-pop catalog: what's been heard, what fits, and what comes next.
Sprout № 03
The Lunar Reel
The public face of a private K-pop scoring system: weekly comeback predictions, rendered as a short video series.
Sprout № 04
The Season Ledger
A personal Drag Race judging app: score each episode as you watch, and see queens rank across seasons through a recency-weighted taste model.