clinical history — behavioral compounds
Behavioral scientist who codes. Category of one.
| Domain | What they build | What they miss |
|---|---|---|
| Engineers | Features, systems, APIs | Why users adopt, avoid, or abandon |
| Designers | Interfaces, flows, prototypes | The psychological mechanism underneath |
| Behavioral Scientists | Research, papers, frameworks | Production code that ships |
The gap: Engineers ship features without understanding the psychology. Designers polish surfaces without measuring the mechanism. Behavioral scientists publish papers without writing a line of production code.
I sit in the intersection of all three. I design the behavioral system, build the native product, and ship it.
I build products where the psychology is part of the architecture — not a layer of polish applied after the feature works, but the reason the feature works at all.
When a team says "the feature works, but users don't adopt it," I find the same underlying failure: the model of the user is too shallow. The interface assumes rational behavior where the real driver is anxiety, ambiguity, status, avoidance, reward anticipation, or social risk.
I diagnose that layer and rebuild it directly into the product. Native iOS apps. Procedural canvas rendering. Web platforms. Engagement systems. Whatever the idea requires.
Your engineering candidates can ship the feature. They can't tell you why users bounce off it. They'll build exactly what the spec says and then watch adoption flatline because the spec modeled a rational user that doesn't exist.
Your design candidates can make it feel right. They can't instrument the feeling. They'll polish the surface and then have no answer when the A/B test shows the polished version performs the same as the rough one, because the problem was never visual.
Your behavioral science candidates can explain the mechanism. They can't build the product. They'll write a research brief that's correct and comprehensive and then hand it to an engineer who strips the psychology during implementation because it wasn't expressed in code.
I am the bridge. I design the behavioral system, write the production code, and ship the product — with the psychology intact, not as a PDF attachment.
50 repos in 2 weeks. 14 casino games in a day. A procedural open-world game with emotional creature AI. A computational persuasion research system. A hormone tracker disguised as a weather app. 27 reusable behavioral design skills.
The throughline: every single thing I build has a psychological mechanism at its core — not as decoration, but as the architecture. The mechanism is why the product works. Remove it and you have a todo app.
| Factor | What it means |
|---|---|
| Speed | 50 repos in 2 weeks. 14 games in a day. I ship in days, not sprints. |
| Range | Native iOS, React web, Canvas games, Python ML, CLI tools, design systems. Whatever the idea requires. |
| Behavioral depth | Every product has a DOSE framework analysis, engagement architecture, and retention mechanics baked in. Not added later. |
| Iteration | 4 generations of conversation training apps. 2 versions of Jeju Island. I refine, not just ship. |
| Systems thinking | 27 codified skills that produce consistent behavioral design. This isn't intuition — it's a system. |
| Taste | Procedural Canvas art. Spring physics. Radial gradients. Specular highlights. I care about how it feels, not just how it works. |