George Railean - UX/UI Designer

What if governance could finally look the way it works?

For decades, public policy has been locked in formats almost no one can hold in their head at once. Hundred-page reports that twenty people open and ten finish. The structure of a government exists on paper, but nobody can see it.

Control AI set out to make it visible. A dark palette. Institutions rendered as nodes, connected by lines that reveal relationships no document ever showed. Underneath, graph analysis and AI-assisted policy mapping surface the connections that would otherwise stay buried across legislation and institutional structure. The Constitution sits where everything converges, which is exactly where it belongs.

This is not an interactive report. It is a working argument that complexity deserves form. A policy analyst reads the same screen as the designer who built it, because the percentages, the connections, and the gaps speak in one visual language.

The alignment scores measure how closely each institution and its legislation track the Constitution's objectives. Those numbers existed before. 43 percent. 25 percent. 14 percent. They sat in spreadsheets without weight. Here they have position and consequence. Statistics became signals.

Control AI does not change the laws. It changes how we see them. Sometimes that is the first step toward changing them.

What if a law wasn't just a document?
What if you could see it - every entity it touches, every regulation it's connected to, every complaint it generates, every percentage point of ex*****on it gains or loses month over month?
That's the question that shaped this project.
Working on the Control AI Policy Platform pushed me to rethink something I thought I understood: how people in power make sense of complexity. Turns out, they often don't. Not because they lack intelligence, but because the tools they use weren't designed for the scale of what they're dealing with.
Hundreds of federal laws. Thousands of relationships between them. Millions of social media signals. Real sentiment from real people, buried somewhere between 24% negative and 44% positive - and someone has to make a decision based on that.
We tried to make that legible.
Not through another table. Not through another report. But through something that feels closer to how the mind actually works - spatial, connected, alive.
Some problems are hard because the technology isn't there yet. This one was hard because the problem itself kept growing the closer we looked at it. Every answer revealed another layer. Every screen we designed had to carry more weight than it looked like it could.
And then, at some point, it did.

What does an entire legislative system look like on a single screen?
That was the question behind every design decision in the Control AI Policy Platform.
We did not want to build another dashboard. We wanted to build a tool where someone managing public policy could see the system they work within, not just read about it.
The Constitution sits at the center. That is not a visual metaphor. It is information architecture.
Policy domains orbit around it. Each one is a cluster of typed nodes: Entities, Legislation, Services, Regulations. Every domain carries its own coverage percentage, and the radar at the edge rolls those up into a system-level view. The timeline on the left scrubs through legislative history. The AI layer at the bottom answers questions in plain language. New simulations can be modeled directly on the same canvas.
Dense legislative data, presented as a workspace instead of a database.
Governance made visible.