Please don't hesitate to reach out. Also, feel free to check out my other links.
TL;DR: Gemini, Google's AI assistant, is genuinely useful in a product design workflow — especially because it's strongly multimodal, so you can hand it screenshots and mockups, not just text. Use it to analyze visual designs, critique hierarchy and accessibility, generate and refine UX copy, and accelerate research. Below is the practical workflow I use, with the same principle that applies to any AI tool: treat output as a draft and apply your own judgment on top.
Gemini is a family of AI models from Google, accessible through the Gemini app and Google AI Studio. For designers, its standout trait is strong multimodal understanding — you can upload an image of a screen and ask it to reason about what it sees, which is a natural fit for visual design work.
Upload a screenshot and ask Gemini to describe the visual hierarchy, identify what draws the eye first, and flag where attention gets lost. This is a fast way to sanity-check a design before you ship it. Prompt: “Look at this screen. What does the eye see first, second, third — and is that the order I want for conversion?”
Ask Gemini to review a design for obvious accessibility issues — contrast, text size, tap-target spacing, unclear labels. It won't replace a real audit, but it catches the obvious problems early when they're cheap to fix.
Like any strong language model, Gemini is useful for microcopy — headlines, button labels, onboarding, empty states. Give it the user's context and ask for several options at different tones, then refine the best one.
Use Gemini to summarize competitor flows, cluster user feedback into themes, or draft research questions. Its integration with Google's ecosystem makes it convenient for pulling together and organizing information quickly.
For early-stage exploration, Gemini is good for breadth — many rough directions fast. Use it to escape the blank canvas, then switch to deliberate craft once you've picked a direction worth pursuing.
Can Gemini analyze images and screenshots? Yes — it's multimodal, so uploading a mockup and asking it to reason about the visual design is one of its most useful features for designers.
Can Gemini replace a designer? No. It accelerates production and analysis, but deciding what to build and whether it serves the business and user is human judgment.
Gemini or another AI tool for design? Gemini is especially strong for multimodal, image-based tasks. Many designers use more than one tool, each for the stage it's best at.
Carlos Lastres is an Apple Design Award–winning product designer and software engineer in Tokyo who works hands-on with AI tools to design conversion-focused products.