This is a new module added to the Google UX Certificate only in 2024+, which integrates Gemini / NotebookLM / Career Dreamer these Google AI tools into the entire UX workflow and job search process. This course uses Google's own tools, but the methodology (especially the T-C-R-E-I prompt framework) is universal, and it works the same when replaced with Claude/GPT.
Significance for PMs: Among the 7 wikis, this course is the freshest and most directly copyable for you — it breaks down "using AI for product work" into specific scenarios + prompt templates. Among them, NotebookLM(document-grounded Q&A) you already have the notebooklm skill on your local machine(local reference materials), and what is provided here are the specific ways to "use it as a project knowledge base / interview preparation library".
1. T-C-R-E-I Prompt Framework (Core, Reused Throughout)
The 5 elements of writing effective Prompts, mnemonic: "Thoughtfully Create Really Excellent Inputs":
| Letter | Element | How to Write |
|---|---|---|
| T | Task | Specify the AI's role and Persona(e.g., "UX content designer with 15 years of experience") + Output Format(list/table/paragraph) |
| C | Context | Target users, product, scenario, style preference (e.g., "The user is 29 years old, loves winter sports, and just switched from snowboarding to skiing") |
| R | References | (Optional) Provide similar examples / styles you want to imitate, so the result better fits your needs |
| E | Evaluate | Check whether the output meets expectations; if not satisfied, first review whether the prompt is specific enough |
| I | Iterate | Revise the prompt, add details, change the format, and repeat until satisfied |
4 Practical Tips to Improve Prompt Effectiveness: ① Clearly state what you are not satisfied with (add clearer guidance) ② Provide examples (desired format/style) ③ Break complex sentences into several simple sentences ④ Tell the AI which outputs you like + why, and let it generate again based on your preferences.
This framework is consistent with the prompt philosophy in your KB's newtype · Prompt / Context Engineering in Practice / Wang Kai's Complete Claude Code Terminal Workflow Methodology; T-C-R-E-I is a lightweight version for non-technical UX practitioners.
2. 7 Scenarios Where Gemini Assists UX
| Scenario | How to Use | Key Points |
|---|---|---|
| Generate Personas | Paste the user research summary + "Give me 4 typical personas, each with a brief description" | Saves organization time and speeds up brainstorming; however, the AI draft must be reviewed and revised by the team, do not over-rely on it |
| Simplify Design Specifications | "According to m3.material.io, should the two buttons on a mobile dialog be aligned left or right? Should they be filled or text-styled?" | Acts like a 24/7 online design consultant that understands jargon (CTA/Material); professional judgment still relies on humans |
| Diverge in Design Iteration | "I am a UX designer working on a language learning app, give iteration suggestions for the following draft ideas" + paste existing ideas | Make full use of Context+References, so AI expands on your basis rather than repeating; it can discover edge cases |
| Improve Stakeholder Communication | "Rewrite this UX jargon-heavy sentence into several jargon-free, concise versions" | Remove jargon — translate phrases like "iterative affinity diagram analysis to identify personas…" into plain language, categorized by "concise version/detailed version" |
| Write Portfolio Copy/Case Studies | "I am a UX designer, write a first-person case study opening based on the following project information" | AI is good at drafting, you are responsible for adding personality; it will mark the parts you need to fill in (App name/target users) |
| Generate Realistic Placeholder Copy | Instead of using "Lorem Ipsum" in design drafts, let AI generate example content close to real scenarios | Makes high-fidelity mockups more realistic |
| Analyze User Research Data | Let AI summarize research findings, identify behavioral patterns, and categorize user types | Humans still need to verify data accuracy |
Iron Rule (Instructor Paolo, Google Maps UX Director): AI is an acceleration tool, not a replacement for thinking and decision-making. "Try it bravely, there is almost zero risk, many tools are free; build the habit of using it, understand how it works, and adapt to industry evolution." Final polishing, evaluation, and implementation still rely on humans.
3. NotebookLM: Document-Grounded Project + Interview Assistant
NotebookLM = Google's AI research assistant,which answers only using the documents you upload(it does not cite external uncertain content, and answers are marked with sources). The fundamental difference from internet-connected AI:information is credible and traceable.
Two Major Uses:
- Project Knowledge Base(Instructor Ben, Google Cloud PM): When collaborating across teams, materials are scattered and emails are long — build a team knowledge base. Three steps: ① Upload documents (product specifications/design plans/timelines) ② Teach colleagues how to ask questions ("What is this project supposed to do?" "What is the next stage deadline?") ③ Get answers with sources. Scenarios: new hire onboarding/writing meeting minutes/listing agendas/finding materials/making FAQs.Note: Documents will not be used to train AI (confidentiality), the more accurate the documents, the more accurate the answers, and it is suitable for stable documents that do not change often.
- Interview Preparation Library: Upload the job JD + company materials + resume + industry reports → ① Summarize key information (quickly grasp company background) ② Generate interview study materials (common questions/answer frameworks/skill guides) ③Generate an audio overview(listen and review during commute).
Your local notebooklm skill uses the teng-lin pip version (full API); there is a dedicated setup recipe for logging in from within the firewall, see wiki-local-env-layout.
4. Full AI-Assisted Job Search Process
Starting Point: Transferable Skills + Career Identity Statement + Career Dreamer
- Transferable Skills(Instructor Mike): Identify the most valuable skills from your past experience that can be carried over to the new position (especially important for career changers entering UX).
- Career Identity Statement: A one-sentence positioning of who you are and what you can bring — the anchor for subsequent resume and cover letter revisions.
- Career Dreamer(Google tool): Use AI to start job search exploration, helping career changers find direction.
AI Resume Revision (2 Steps)
- Build a Basic Resume: Showcase core skills,quantify achievements(e.g., "improved satisfaction by 25%"), use clear formatting, and make it usable as a customization template. Feed your old resume + career identity statement to Gemini: "Give me specific, actionable revision suggestions to make me more competitive for position X." → AI suggests rewriting with transferable skills, adding more quantification, and provides sample text.
- Tailor to the Position: Give the target JD + basic resume together to Gemini: "Based on this job description, suggest how I should adjust my resume to become the ideal candidate." → AI matches keywords and suggests additions/deletions (e.g., adding A/B testing experience). Name each version (
Sage_Resume_Google_PM.pdf) for archiving.Always stay truthful.
Supporting: Write a personalized cover letter, optimize LinkedIn/online presence (from recruiters Olga/Antonio's perspective: online presence has both opportunities and pitfalls).
Interview: STAR + Formats + Gemini Live Practice
- Interview best practices + formats: Familiarize yourself with different interview formats (phone/video/in-person/portfolio presentation).
- STAR Method for answering behavioral questions:S Situation /T Task /A Action /R Result — structure the telling of your experience.
- Gemini Live Mock Interview: Practice answering questions in real time with AI voice. NotebookLM generates company FAQs + answer frameworks.
- Prepare thoughtful questions for the interviewer(questions to ask the interviewer), handle tricky questions (instructor Julia's confident response approach),follow up after the interview(thank-you note),respond to offers.
Job Search Management: Networking + Resilience
- Networking(Instructor A.J./Module 3): One of the most effective channels for job searching; proactively build and use networks to strengthen your job search.
- Resilience: Job searching is a long run, rejection is normal, maintain a good mindset (echoing the entry stories of Lisa/Juan/Michael in Google UX Certificate / UX Design Fundamentals — none of them had a smooth start).
Key Terms Glossary
| English | Chinese | One-Sentence Explanation |
|---|---|---|
| T-C-R-E-I | 5 Elements of a Prompt | Task/Context/References/Evaluate/Iterate |
| Persona (in prompt) | Persona Setting | Have AI take on a specific expert identity |
| NotebookLM | — | AI that answers only using documents you upload (traceable) |
| Career Identity Statement | Career Identity Statement | A one-sentence positioning of what you can bring |
| Transferable Skills | Transferable Skills | Abilities from past experience that can be carried over to a new position |
| Career Dreamer | — | Google AI job search exploration tool |
| STAR Method | STAR Method | Situation/Task/Action/Result, used to answer behavioral interview questions |
| Gemini Live | — | AI voice real-time mock interview |
<!-- Video gap: This topic includes instructor demo videos such as Generate personas / Simplify frameworks / Improve stakeholder comm / Use AI iteration / portfolio copy — the text already covers the methods and prompts; a small number of raw instructor interview videos (Paolo et al.) have not been transcribed, and their content has been incorporated. -->