AI Lab

I don't just work faster with AI. I build the tools that let my whole design org do it too.

Everyone at Procore can spin up an AI prototype now. What actually moves the needle for a design org is building the infrastructure that lets everyone else do it well, too.

Below is a small library of AI skills I built that encode Procore's design system and workflows once, so any designer, PM, or engineer can reuse them — plus the AI-assisted work I shipped using them.

AI Lab

4 entries

Design Orchestrator — a shared pipeline for turning Figma into working prototypes

Design Systems Tooling

View case study

Design Orchestrator — a shared pipeline for turning Figma into working prototypes

Org-level AI tooling — Procore design system, composed once

2026

A reusable AI skill that composes Procore's design-system knowledge — the core-react component library, data-table and formulator patterns — into a repeatable pipeline for turning a Figma file into a working HTML or React prototype. Built so any designer can point it at their own designs, not just a personal script.

AI ToolingDesign SystemsPrototypingProcore

View case study

Turning a feature request into a five-document pipeline an agent can pick back up

Procore Document Management — Drag-and-drop file uploads

2026

Procore Document Management needed native drag-and-drop uploads — a pattern competitors already had. Rather than write a design brief by hand after the code shipped, I used it as the proving ground for an AI-assisted planning pipeline: product brief, feature briefs, design brief, and engineering handoff, chained through a resumable checkpoint file.

AI ToolingProcess DesignProcoreDocument Management

View case study

An interactive résumé that became an editable Figma file — no re-drawing required

Personal project — Cursor Canvas to Figma, via MCP

2026

Built a dynamically-styled, interactive résumé as a live Cursor Canvas artifact, then used Figma's MCP tooling to translate it into a fully editable Figma file — real components and layers, not a flattened screenshot pasted into a frame.

Personal ProjectAI ToolingFigma MCP

Two AI-built prototypes, one real decision: hand-rolled table or Smart Grid?

Technical De-risking

View case study

Two AI-built prototypes, one real decision: hand-rolled table or Smart Grid?

Procore Document Management — Documents table refresh

2026

Built parallel AI-assisted prototypes of the same Documents table — one hand-rolled, one on Procore's real Smart Grid (AG-Grid-based) component — to make the trade-offs concrete before committing real engineering time to either approach.

AI ToolingTechnical EvaluationProcoreData Tables