Rod Del Castillo Process + Workflow

How I build with people, systems, and AI.

Senior-to-principal IC product design, rooted in whiteboards and product judgment, with AI used as leverage for prototyping, automation, QA, and zero-to-one build work.

Practice Pillars

Clarify Before Making

I start by separating signal from noise: what the product needs to do, who it is really for, what decisions are still open, and what can be cut without weakening the outcome.

Stay Close to the Material

I move between sketches, Figma, prototypes, code, content, and motion so the product is tested in the form people will actually use, not only as a polished static screen.

Use AI Without Losing the Thread

AI helps me synthesize, explore, build, check, document, and automate. The point is not to make the work feel artificial; it is to protect more time for the judgment calls that still need a person in the chair.

Process

Frame

Stakeholder conversations, product audits, whiteboarding, rough notes, and a clear read on the real decision the team needs to make.

Structure

Flows, IA, states, constraints, and system boundaries before the work turns into a pile of disconnected screens.

Prototype

Fast models of the experience in the right fidelity: sketches for thinking, Figma for alignment, code when behavior needs to be felt.

Refine

Build-ready UI, interaction details, responsive behavior, content, and design-system decisions that can survive handoff.

Ship

Embedded support through implementation, QA, spec cleanup, automation opportunities, and the last-mile polish that keeps the idea intact.

AI as Leverage

Research & Synthesis

I use LLMs to organize interviews, notes, competitive audits, support tickets, and messy source material into themes I can interrogate. The tools do the first sort; I still make the call.

Concept Exploration

I generate a wide field of references, language, interface directions, and motion ideas quickly, then edit hard. AI is useful for range; the product still needs taste.

Zero-to-One Prototyping

I can go past the deck and into something testable: a clickable model, a coded prototype, a local tool, a landing flow, or a rough product surface that makes the idea easier to judge.

Workflow Automation

When the same task keeps coming back, I look for a system: generated docs, structured QA passes, content cleanup, asset prep, internal utilities, or small agents that remove drag from the team.

Design QA & Handoff

I use AI-assisted checks to compare intent against implementation, catch copy and state issues, tighten documentation, and keep design decisions visible after the file leaves Figma.

Learn & Iterate

After launch, I use the same loop to summarize feedback, inspect friction, compare product behavior against the original bet, and decide what deserves another pass.