Will's AI-Assisted UX Workflow Overview (Full A-Discover→F-Deliver Phase Skeleton + Navigation for 12 Sister Wiki Pages)

This is the "table of contents / entry page" for the entire will-ai-note will-ai-note series. One-sentence positioning:This set of notes teaches you to "do UX with AI", not "how to do UX"(for the latter, see the UX craft series under the Google UX Certificate / UX Design Fundamentals Google UX Certificate / UX Design Fundamentals).

What this workflow is: A pipeline "from discovering problems to delivering and presenting"

Will breaks down a complete UX design project, from "I notice a problem" all the way to "I can explain it clearly in interviews / reviews / my portfolio", into a sequential pipeline. Core idea:

String together the full process from problem discovery to delivery and presentation to form a reusable thinking framework.

Main thread (English = Will's original naming, Chinese = plain language explanation):

A  Discover & Research   发现问题 & 用户研究
B  Define & Insight      定义问题 & 提炼洞察
C  Ideate               方案构思
D  Experience Design     体验设计(界面 / 状态 / 动效)
E  Validate             验证(A/B 对比 + 指标)
F  Deliver              交付(讲故事 / 面试稿 / 作品集)

Why this order (Why): This line is essentially a cycle of "first converge, then diverge, then converge again" —

1. A first lays out the full background: Without clarifying the business context and user scenarios, everything that follows is rootless.
2. B converges vague problems into verifiable hypotheses and insights: Converging from "I feel this is not usable" to "My bet is X, and the verification method is Y".
3. C diverges into strategies, information architecture, feature logic, and solution documents on top of insights: This is "figuring out clearly what to build".
4. D turns solutions into concrete interfaces, states, motion effects, and high-fidelity prototypes: This is "what it looks like when built".
5. E validates with comparative data and metrics: Prove the redesign is actually better, and that the conclusion is logically self-consistent.
6. F packages the entire journey into a narrative that can be presented externally: Interview verbatim scripts + portfolio page layouts.

Life analogy: It is like writing a persuasive paper. A is selecting a topic and researching background, B is establishing a thesis and finding evidence, C is building the paper outline, D is writing out and polishing every paragraph, E is running experiments to get data for proof, F is how to present it during defense.If the order is wrong (for example, starting to write the main text before establishing the thesis), you will definitely have to rework later.

What role AI plays in each phase (the real value of this set of notes)

Everyone knows the ordinary UX process. The differentiator of Will's system is:Every node is marked with "how AI intervenes", so that you "both know what to do, and immediately learn how to do it with AI". AI's role changes with the phase:

A key principle running through the whole process (Why it matters): Research data, comparative evidence, and validation materials generated by AI should first be used via "expanded generation" to fill gaps you cannot get in real projects(especially in personal portfolio scenarios where there are no real users);once you obtain real data and feedback, go back and replace the simulated materials at the corresponding nodes. This avoids "deceiving yourself with AI-made-up data" — it is only scaffolding, not the conclusion.

Phase navigation table: What to do → which wiki to go to for depth

Each phase is broken into numbered sub-steps (a-1, b-1…) in Will's original notes; below is the skeleton + corresponding in-depth wiki:

Note: E-Validate is a separate block in Will's original diagram, but its "data supplementation + visualization + metric proof" is essentially evidence preparation serving F delivery, so this series treats it as a pre-step for F.

Tool section navigation: Specific "how to use AI" skills supporting each phase

Phase wikis cover "at which step to use AI", while the tool section covers "exactly which tool to use and how to tune it". These are cross-phase general skills:

Relationship with the google-ux craft series: Walking on two legs

It is easy to confuse these two sets of notes; remember one sentence to tell them apart:

• Google UX Certificate / UX Design Fundamentals series = "how to do UX"(UX craft): the core design skills themselves, such as user research, information architecture, usability, and visual specifications.. This is the foundation you should already have before the AI era.
• This will-ai-note series = "doing UX with AI": On the premise that you already understand UX,let AI speed up every step, fill gaps, and produce materials. This is an amplifier.

Why view them separately (Why): AI is an amplifier, not a replacement.The more solid your underlying UX craft, the more reliable what AI amplifies will be; if your UX skill is 0, what AI helps you efficiently produce is also 0. So the ideal order is to first lay a solid foundation with Google UX Certificate / UX Design Fundamentals, then use this series to maximize efficiency.

One-page cheat sheet (for when you are in a hurry)

• What this set is: An A→F AI-assisted UX pipeline that connects the full process from problem discovery to delivery and presentation.
• How to use it: Enter the corresponding wiki from the phase you are currently in (if you are at A finding problems, read AI-Assisted UX · Discovery & Research Phase (Prompts and workflows for using AI to define business/experience goals, MVP metrics, problem positioning, user research reports, and background infographics); if you need to deliver a portfolio, read AI-Assisted UX · Delivery Phase (Using AI for layout design, layout-to-code in Figma, solution reviews, demonstrating AI value + will's custom UX metric breakdown skill)), and flip to the tool section when you need specific tool skills.
• Core principle: Research / validation materials generated by AI are scaffolding; you must backfill and replace them once you get real data.
• Do not mix them up: This series = doing UX with AI;Google UX Certificate / UX Design Fundamentals = how to do UX. Two legs: foundation first, then amplification.

English Summary (bilingual)

What this is. The entry/index page for the will-ai-note series — Will's "AI-empowered UX workflow," a reusable pipeline running from discovering a problem to delivering the narrative.

The backbone. A Discover & Research → B Define & Insight → C Ideate → D Experience Design → (E Validate) → F Deliver. The logic is converge-diverge-converge: A frames context, B converges a vague problem into testable hypotheses & insights, C diverges into strategy/IA/flows/spec docs, D realizes concrete UI/states/motion/prototypes, E proves it with A/B data & metrics, F packages the whole journey into an interview script + portfolio.

AI's role shifts by stage. Research assistant (A) → simulated researcher (B, generate stand-in data) → structurer (C) → design executor / image gen (D) → comparison visualizer (E) → narrative & layout helper (F). Key rule: AI-generated research/validation material is scaffolding to fill real-world gaps; once you have real data, replace the simulated nodes.

Navigation. Stage wikis: AI-Assisted UX · Discovery & Research Phase (Prompts and workflows for using AI to define business/experience goals, MVP metrics, problem positioning, user research reports, and background infographics) · AI-Assisted UX · Definition & Insight Phase (Prompts and workflows for using AI to extract insights, visualize user research, select chart types, interpret eye-tracking, and score solutions) · AI-Assisted UX · Ideation Phase (Prompts and workflows for using AI to produce project decision reports, competitive analysis, innovative product generation, and user flow wireframes) · AI-Assisted UX · Experience/UI Design Phase (Prompts and workflows for using AI to generate polished UI, design systems, color schemes, interaction states, microcopy, UI-to-code conversion in Figma, and responsive design) · AI-Assisted UX · Delivery Phase (Using AI for layout design, layout-to-code conversion in Figma, solution reviews, demonstrating AI value + will's custom UX metric breakdown skill). Tooling wikis: Highlights of Google Prompting Essentials Course (Prompt frameworks / 4 iteration methods / prompt chaining / common scenario playbooks) · AI-Driven Product Design / Vibe Coding Process (One-sentence requirement → interrogation → PRD → DRD + coding prompts → Figma Make demo → launch → visual iteration) · Figma Make Usage Manual (Operations and use cases for basics/templates/applying design library styles/adding Guidelines/open-source components/adding backend/publishing) · Codex + Figma Toolchain (Steps and skills for installing Figma skill/MCP, generating chart component library slides, converting prototypes to Figma, and automatically replacing SVG icons) · AI Visual and Motion Design Techniques (Prompts and tools for image upscaling and sharpening/3D icon and IP design/KV key visuals/stylized illustrations/visual remix reconstruction) · Installing and Using Skills on Claude.ai (Step-by-step operations + when to use) · Will's Core Principles for Using AI Bots + Common UX Interview Questions + Prompts for Creative Project Explanations.

Relation to the craft series. This series = doing UX with AI; Google UX Certificate / UX Design Fundamentals = how to do UX (the craft). AI is an amplifier, not a replacement — lay the craft foundation first, then turn the AI amplifier up.