Show HN: I used Gemini 3 Pro to design my landing page(lingoku.ai)

Show HN: I used Gemini 3 Pro to design my landing page

https://lingoku.ai/learn-japanese

I'm a backend engineer with almost zero design experience. I experimented with a “Dual AI” workflow to brute-force a professional landing page.

What I built

  A landing page for Lingoku (language learning extension): https://lingoku.ai/en/learn-japanese

The workflow (short)

  - The Hands (Figma Make): auto-generated raw UI drafts from my feature list.

  - The Brain (Gemini 3 Pro): I fed screenshots and asked it to act as a Senior Designer — critique colors, visual hierarchy, clarity, and trust signals.

  - The Loop: I applied Gemini’s specific feedback in Figma and repeated — dozens of “roast & fix” iterations.

What I want from HN

  Be blunt — is this page actually professional? If not, what are the top 3 things you’d change immediately?
  Also: can you suggest a step-by-step AI-assisted workflow I can follow to improve this (tools, prompts, and order of operations)?

Context / constraints

  I'm a solo backend dev; I want a reproducible process I can run with AI + Figma.

  UX/performance tradeoffs are fine — prioritize visual clarity, trust, and conversion.

Thanks for any critiques, specific edits, or prompt examples — I’ll iterate and share updates.

2 コメント

englishcat7 か月前

Thanks for checking this out — short context: I’m a backend dev, iterated ~30 times using Gemini 3 Pro for critique. If you reply, please tell me:

Does it read/professionally look the part?

Top 3 changes you’d make.

One prompt or tool I should try next.

I’ll apply suggestions and post an update — much appreciated.

frost_dev7 か月前

Hello, I just went through your site and I really love your work. Though I found that you used emojis there...You can use some cool icons that work better.

Also, it is not responsive on mobile...

If you want, I can assist you in fixing some errors