Foundations: enrolling now. Self-paced. $299. Join the list.

About Joinery

Every generation of creatives gets a new tool that threatens to flatten their work. Templates did it. The web did it. AI is doing it faster than any of them. The structural problem is always the same: the tool fills in its own defaults wherever your intent isn't explicit enough. Joinery teaches you to keep your work yours, whether you're working with AI, building a brand on Shopify, writing a book, or governing a design system.

The methodology comes from a classroom, not a lab. In special education, designing around how each student actually processes is federal law. You write a plan for each learner. You observe how they respond. You revise the plan based on what you see, not what you assumed. Ignore that process and you get compliance. Design around it and you get learning.

That same move (read the room, design around what you find) applies in every domain where the gap between intent and execution is the whole problem. Kitchens, printshops, enterprise platforms, recording studios. AI is where it became visible enough to formalize. Six published papers document the methodology. But the skill is older than any of them.

Joinery teaches practitioners to apply a skill they already have. You already read the room. You already read the client, the student, the brief, the audience. The courses show you how to apply that same reading to whatever system you're working with, so the output carries your intent instead of the tool's defaults.

Peter Salvato

Peter Salvato

SVA-trained designer. Taught special education in a self-contained classroom in Sunset Park, Brooklyn: twelve students, every subject, twelve individual plans, no other teacher running the room. Twelve years as product design lead on an enterprise platform through three ownership changes. Before any of that, growing up on construction sites in Brooklyn, watching his father take buildings apart and put them back together.

Every room asked the same question. What does this system actually need, and how do I design around that reality instead of the one I assumed? On a construction site, you read the structure before you cut. In a classroom, you read the student before you write the plan. On a twelve-year-old codebase, you read what drifted before you redesign. In a kitchen feeding four people with incompatible diets, you read constraints before you cook. In a DJ booth, you read the room before you select the next track. Writing a novel, you read the story's pace before you write the next scene.

Twenty-five years of that same move across printshops, classrooms, enterprise platforms, kitchens, recording studios, and prose. AI was where the pattern became clear enough to name. Peter published six papers with DOIs formalizing what he'd been doing instinctively across every other domain. The methodology didn't come from studying AI. It came from decades of reading rooms and designing around what he found. AI was the room where it finally became teachable.

Design Education School of Visual Arts: Christoph Niemann and Nicholas Blechman taught that every mark carries meaning, precision as discipline. James Victore taught that your instinct is the material, fierceness as honesty. The synthesis: communicate with fierce specificity. Brooklyn College (M.A. program in education). Kingsborough Community College (taught intro to graphic design and visual studies). NYC Department of Education (special education, self-contained classroom, Sunset Park).
Design Practice Twenty-five years across graphic design, print production, brand systems, and product design. Twelve years as product design lead in enterprise software. Brand systems for independent businesses. Construction sites, printshops, kitchens, DJ booths, a novel. The same skill applied to every material.
Research Six published papers with DOIs formalizing a cross-domain practice into a teachable methodology. Accommodation Design, Input Inversion, Prosthetic Cognition, Semantic Hierarchy, Lens Extraction, Voice Governance. Citations include Sweller, Cowan, Murdock, Liu et al., Bock, Pennebaker, Flower & Hayes, Bereiter & Scardamalia, Wong et al.
Infrastructure 130+ working AI skills, agents, and voice protocols in daily production use. The methodology is built from the same tooling it teaches.