How I Use AI to Design Better Products
Artificial Intelligence is no longer a novelty in my workflow. It has become a reliable, mission-critical co-pilot—one that helps me move faster without compromising judgment, clarity, or design quality. I don't use AI to replace thinking; I use it to amplify it.
By integrating AI across research, design, systems thinking, and engineering, I've been able to reduce friction, eliminate repetitive work, and focus more deeply on decision-making, user empathy, and execution precision.
AI as a Research Partner, Not a Shortcut
User research is one of the most cognitively demanding parts of product design. I use AI—primarily ChatGPT and Claude—to synthesize qualitative and quantitative inputs: interview notes, product feedback, reviews, internal assumptions, and edge cases.
Instead of asking surface-level questions, I prompt AI to:
- Identify recurring behavioral patterns
- Surface hidden or underserved user needs
- Stress-test assumptions
- Generate alternative personas and edge scenarios
This allows me to approach Figma with clarity. Before a single pixel is placed, I already understand who the user is, what they struggle with, and where the product can fail if not designed carefully. AI becomes a thinking partner that challenges my perspective rather than blindly agreeing with it.
Precision Prompting: Treating AI Like a Senior Teammate
There is a clear difference between casual prompting and professional prompting. I never ask AI to "design a landing page." I treat it the same way I would brief a senior designer or engineer.
A typical prompt might sound like:
"Act as a senior product designer with 10+ years of experience in SaaS. We are designing a modern B2B landing page with a soft but confident visual tone. Use a 12px border radius system, a neutral base palette, and #F26D40 as the primary accent. Typography should prioritize readability and hierarchy. The layout must be responsive, conversion-focused, and engineering-friendly."
By defining constraints, intent, and context, I get outputs that align with real-world design systems—not generic UI concepts. This approach turns AI into a tactical tool rather than a random generator.
Designing Better Flows Through Systems Thinking
One of the most underrated ways AI helps me is in flow validation. I use AI to map empty states, loading and error states, permission and failure scenarios, and alternate paths users may take.
These are often the details that get missed during early design phases but cause real UX friction in production. By having AI actively look for "what's missing," I can design more resilient, production-ready experiences—especially important when working closely with engineers or building myself.
Tooling: How AI Fits Into My Design Stack
AI isn't a single tool—it's an ecosystem:
- ChatGPT & Claude – Research synthesis, flow analysis, system thinking, prompt iteration
- Figma Make – Rapid ideation, layout exploration, and component variations
- Antigravity & Claude – Translating design intent into clean, structured frontend logic
- Gemini – Illustration support, visual ideation, and image generation when custom assets are needed
- Cursor / VS Code Copilot – Bridging design and code efficiently, especially when prototyping or building real features
AI Amplifies Judgment — It Doesn't Replace It
AI doesn't make design decisions for me. It doesn't understand users the way humans do. What it does is remove friction—so I can think more clearly, test ideas faster, and focus on what truly matters.
As a design engineer, my value lies in judgment, empathy, systems thinking, and technical awareness. AI simply supports those strengths. The result is better work, delivered faster, with more intention and fewer blind spots.