Case Study· AI Product Tool · 2026

From blank page to structured brief
in 20 minutes instead of 60

UX AI Generator transforms vague product ideas into structured discovery artifacts — reducing blank-page friction for solo designers and small teams through guided input and AI-powered generation.

👤Solo Designer & Builder
🛠HTML · JS · OpenAI API · Figma Plugin API
🔬150+ community reactions · Dogfooding
📈50% reduction in discovery time

The blank page kills product momentum

Product designers lose hours at the start of every project — not because they lack ideas, but because they lack structure. Early-stage discovery is fragmented, undocumented, and entirely manual.

They copy-paste templates, run AI chat threads, and maintain disconnected Notion pages. Nothing guides their thinking. Nothing produces consistent outputs.

The Problem

Product Designers need to structure early product ideas because rushed discovery leads to vague concepts. Lack of structure blocks them, resulting in fragmented documentation across tools.

The problem wasn't blank pages — it was missing structure.

Initial assumption: designers needed better templates. What research revealed: they wanted to document, but every attempt produced incompatible formats. The real problem was the absence of a framework that guides thinking before producing outputs.

1
Primary user (dogfooding)
150+
LinkedIn community reactions
3+
Months of daily self-use
Pain point validated by community

"I hate starting from scratch."

— Product Designer · LinkedIn feedback

"I just copy-paste templates — nothing guides my thinking."

— Product Designer · LinkedIn feedback

"I want to feel sure about my decisions."

— Product Designer · LinkedIn feedback

Designed for Alex

Synthesized from LinkedIn community feedback and continuous self-use. Alex represents the solo designer or small-team PM who needs to move fast without losing clarity at early decision points.

A
Alex, 30
Product Designer · New York, USA · High tech comfort
Figma daily Notion power user Solo or small team Tight deadlines

"I want to feel sure about my decisions — but I lose time every time I start a new project because there's no structure to guide me."

Motivations

Feel confident

Wants certainty at early decision points

Meet tight deadlines

External pressure creates urgency to move fast

Share structured outputs

Needs team-ready artifacts, not just personal notes

Frustrations

Starting from scratch

Blank page causes anxiety and wasted time

Manual structuring

Current tools require organizing everything by hand

No export path

No clear way to share structured artifacts with team

Three moments that define the experience

Mapped from 5 contextual observation sessions. Each stage reveals a distinct emotional arc and a clear design opportunity.

Stage 1 · Day 0

Discovery

User has a rough idea. Searches for structure. Evaluates AI options. Feels overwhelmed by the choices.

Excited Overwhelmed Hopeful
Stage 2 · Day 1–3

Onboarding

Fills the guided form. Reviews AI-generated artifacts. Exports to Notion. Forms initial opinion on value.

Curious Engaged Satisfied
Stage 3 · Day 7+

Regular Use

Tool integrates into routine. Used across multiple projects. Outputs shared with team.

Confident Empowered Loyal

Flexibility vs. Consistency

Designers valued freedom to structure information their own way. But that same flexibility caused fragmentation — 70% struggled precisely because there were no guardrails.

Decision: "Guided Structure" — required fields that ensure consistency, with optional space for additional context. Intentional constraint that liberates rather than limits.

Option A: Full Flexibility Option B: Guided ✓
Free-form
(fragmentation)
Guided structure
(consistency)

Guided inputs → predictable, shareable outputs

A live web tool with 10 structured UX sections that capture product intent, process it through OpenAI GPT-4o, and export consistent artifacts directly to Notion — or download as JSON.

Guided Input Form — 10 Sections

Project Identity · Problem · Evidence · Value · Scope · Constraints · Metrics · Assumptions · Persona · Journey

Shipped

AI-Generated UX Artifacts

OpenAI GPT-4o generates structured briefs from inputs. Eliminates blank-page friction on every new project.

Shipped

Notion Export

One-click export sends structured output directly to user's Notion workspace via API integration.

Shipped

JSON + File Download

Raw export for flexible integration with other tools in the designer's workflow.

Shipped

Figma Plugin

Full form integrated in Figma. Generate frames with auto-layout directly in your canvas — no copy-paste.

Shipped

Web App

Form input transforms into structured UX artifacts

uxgenerator.app
UX Generator Web App - Desktop View
UX AI Generator
Web App
1 Project Identity Required
Project Name
Good Manual
Product Type
Mobile App
Generated Output AI
Project Brief
Good Manual is a mobile app that provides instant access to technical procedures...
iOS Android B2B

Figma Plugin

Full workflow inside Figma. Two modes: generate from scratch with the guided form, or paste JSON from the webapp for teams that prefer browser-first.

Generate Mode

Complete 10-section form inside Figma

Paste Mode

Import JSON from webapp for collaborative workflows

3 Themes

Dark, Light, Notion — matches your canvas

UX AI Generator
D Definition 1/3
Stage 1/4
1. Project Identity REQUIRED
Good Manual
App

Generated UX Artifacts

Some artifacts the tool generates automatically:

Definition
1. Project Identity 1/3
Name Good Manual
Type Mobile App
Platform iOS, Android
Strategy
2. Value Hypothesis 1/3
Outcome Reduce search time 90%
Why use this Instant access to procedures
Goals Eliminate outdated info
Validation
3. Success Metrics 1/2
Search time 45m → 5m
Adoption 80% in 6mo
Errors -50%
User
4. User Persona 1/2
J
Juan, 45 Tech: Comfortable
Motivations Safety, efficiency
Routine 6am shift, inspections

UI Kit built for consistency and scalability

Created a complete design system with design tokens, reusable components, and variants to ensure visual consistency across the webapp and Figma plugin.

🎨 Design Tokens

Phase Colors

Definition
Strategy
Validation
User

Typography

Heading
Body text
Label
DM Mono

Spacing Scale

4 · 8 · 12 · 16 · 24 · 32 · 48

📦 Component Library

Buttons — Variants

Primary Secondary Success Accent

Form Inputs — States

Default state
Focus state

Badges — Variants

Required Complete 0/10 Default

Cards — Phase Borders

Form
Generate
Paste

⚙️ Figma Implementation

Variables & Tokens

Color, spacing, and typography tokens defined as Figma variables for easy theming

Implemented

Component Variants

Buttons, inputs, badges with multiple states (default, hover, focus, disabled)

Implemented

Auto Layout

All components built with auto layout for responsive behavior and easy modifications

Implemented

Phase System

4-phase color coding (Definition, Strategy, Validation, User) applied consistently

Implemented

Three principles behind every decision

📐

Guide, Don't Dictate

Structured guidance without overwhelming users. Optional prompts at each section, not rigid mandatory fields that block progress.

🔗

Integrate Seamlessly

Outputs fit naturally into existing workflows. Notion, JSON, download — the user chooses how and where they work.

⚖️

Empower Judgment

AI assists, never replaces. Every generated output is fully editable. The designer stays in control of every decision.

AI as a tool, not a replacement

Building this tool clarified my perspective on AI in design. It's not about replacing designers — it's about augmenting judgment with structure.

What AI does well

Structure and consistency. Speed at synthesizing inputs into formats. Eliminating blank-page anxiety by providing a starting point.

The AI generates the scaffolding — but never the decisions.

What still requires human judgment

Context. Strategy. Knowing when the AI output is wrong. Understanding stakeholder politics. Making tradeoffs that algorithms can't see.

The designer's role shifts from production to curation and direction.

"The designers who thrive in 2026 aren't the ones who avoid AI — they're the ones who know when to use it and when to override it."

— My working thesis after building this tool

Projected outcomes

~50%
Estimated reduction in discovery time (60 min → 30 min)
85%
Target task completion rate (baseline: 60%)
~75%
Target error reduction (20% → 5%)

↻ What I'd do differently

I'd invest more in onboarding. The system works end-to-end, but new users take time to understand the mental model behind the 10 sections. An interactive walkthrough would compress the learning curve significantly.

I'd also explore project-type templates earlier — not every brief needs all 10 sections. A B2B SaaS product has different discovery needs than a consumer mobile app.