Career reality; case-studies-as-stories; the extraction interview; the Generic Detector; disclosure.
All tracks — 4 hours, drop-in friendly
Jobs, Portfolios, Freelance, and Creative Career Leverage
Use AI to make your portfolio, CV, and outreach sharper and more yours — not more generic — and leave with a job-search prompt pack you'll reuse for years.
01This Week at a Glance
You will make: one rewritten case-study summary (or CV section) in your real voice + a job-search prompt pack.
Job-search systems — research, role matching, outreach, interview prep, freelance admin.
The same project written two ways: generic vs. extraction-interview.
Rewrite + prompt pack; rewrites read cold — does it sound like a person?
The week in one line
AI is your editor and interviewer, never your ghostwriter. The moment every applicant has AI write their cover letter is the moment AI-voice becomes a disqualifier. Generic is free now, so generic is worthless — the branding lesson, aimed at your career.
02Key Concepts
1 · The career reality, without the panic
Some production tasks are automating; the roles that remain reward judgment, taste, accountability, and critical AI use — the spine of this course. The CV reframe: don't write "good at AI tools" (everyone does). Write what's rare and true — you can direct AI, verify its output, and know when to refuse.
2 · Case studies are stories, not galleries
A case study is a narrative: client/brief → problem → process (your decisions, iterations, rejected directions) → outcome. Recruiters read to imagine working with you; the path is the juicy part. AI can't write your path — it wasn't there — but it's excellent at structuring one you feed it.
3 · The extraction interview (the core technique)
Don't ask AI to write your case study. Make it interview you, one question at a time, hunting for the constraint that hurt, the decision you'd defend, the thing you cut, what the client actually said, what you'd do differently. Then: "write using ONLY my answers, my phrasing where possible." Its questions, your answers, your authorship.
4 · The Generic Detector
AI-voice has tells — "elevate," "seamless," "passionate about," "stunning," "user-centric," "I'm excited to" — recognizable to every recruiter because every AI writes them. The skill: spot them (in AI's output and your own drafts) and replace each with something only you could have written.
Watch out
AI also inflates — claiming results you didn't measure, a role bigger than you had. Deflate it. A modest true claim beats an impressive unverifiable one; recruiters check, and getting caught inflating ends candidacies.
5 · Disclosure as a differentiator
When AI helped with a project, specific disclosure reads as confidence and process maturity. "Textures generated, then repainted by hand; concept and art direction mine" tells a recruiter you know exactly where your value sits. Vague "AI-assisted" invites the worst assumption. Start from the disclosure templates.
6 · Job-search systems (AI as research + admin engine)
- Company/role research: what does this studio make, who are their clients (verify against the real site).
- Role matching: job ad + your skills → where you fit, where the gap is.
- Outreach: you write the personal hook; AI structures around it; you de-generic it.
- Interview prep: "act as the hiring manager and ask me the 8 questions they'd actually ask."
- Freelance admin: lead trackers, follow-ups, scope questions, rate research.
03This Week's Words
Case studyThe portfolio write-up: problem, process, decisions, outcome.
Portfolio disclosureHonest, specific notes on what AI did. Done well, a differentiator.
Extraction interviewMaking AI interview you to surface the specifics only you know.
Generic detectorSpotting and replacing AI-voice stock phrases.
Process documentationIterations, rejected directions, decisions kept — authorship evidence.
Role matchingComparing a job ad to your real skills; finding fit and gaps honestly.
Impact-firstLeading with outcome, not task.
Outreach hookThe specific personal reason you want this place — the part AI can't write.
AI-voice inflationThe model's habit of overclaiming results and roles. Deflate it.
Career boundaryWhat you'll disclose, and where your value genuinely sits.
04Demo: Case Study Rewrite — Generic to Yours
I made this, so can you
One real project written two ways — proving AI is useless as a ghostwriter and powerful as an interviewer.
Prepared demo artifact
Instructor backup / finished example: case-study-diff.html. Use it if the live tool run fails, or to reveal the finished structure after critique.
- Generate the generic version (knowingly): "write a portfolio case study for this project." Read it flat; the room calls the stock phrases; highlight them live.
- The extraction interview: "don't write yet — interview me," hunting for the constraint that hurt, the decision you'd defend, the thing you cut. Answer with specifics.
- The honest rewrite: "write 150 words using ONLY my answers, my phrasing, no adjectives I didn't earn — end with a disclosure line."
- Side by side: the room names the differences — specifics, stakes, voice, the disclosure line as a trust feature.
The planned failure
Version 2 still smuggles in 1–2 stock phrases. Find them, replace by hand, on screen — the Generic Detector is never AI's job to finish.
05Step-by-Step Workflows
The case study rewrite (any project)
- Generate the generic version knowingly — your "before" and stock-phrase target.
- Run the extraction interview — answer honestly, with specifics and stakes.
- Rewrite from your answers only — your phrasing, no unearned adjectives, a disclosure line if AI touched the project.
- Run the Generic Detector — replace every surviving stock phrase by hand.
- Deflate — keep only what's true and defensible in an interview.
- The 5-metre test — would a skimming recruiter learn something only-you in two sentences?
Build your job-search prompt pack
- Save the five system prompts into one doc.
- Customize each with your track, strengths, non-negotiables.
- Use them as templates per application — reusable for years.
- Add moves that work — the pack is a living career tool.
06Prompt Recipes
The extraction interview (the core)I'm writing a portfolio case study for this project: [2-line description]. Don't write anything yet. Interview me, one question at a time, max 6 questions, hunting for: the constraint that hurt, the decision I'd defend in a fight, the thing I cut that I loved, what the client/audience actually said, and what I'd do differently. After my answers, write a 150-word case study using ONLY my material, my phrasing where possible, no adjectives I didn't earn. End with one line disclosing how AI was used in the project (if at all).
The Generic DetectorRead this portfolio/CV text and flag every phrase that could appear in any creative's portfolio — the AI-voice tells (e.g., "passionate about", "seamless", "elevate", "stunning", "user-centric"). For each, ask me a question that would help me replace it with something only I could have written. Don't rewrite it for me.
The deflatorHere is my case study / CV section: [paste]. Challenge every claim about results, my role, and impact. Which are measurable, which are vague, which might be inflated? Rewrite only the overclaims as modest, defensible, true statements I could back up in an interview.
The role matcherHere is a job ad: [paste]. Here are my real skills and projects: [paste/list]. Where do I genuinely fit this role, where's the honest gap, and what 2 things should I learn or show to close it? Don't flatter me — I need the truth before I apply.
The interview rehearsalAct as the hiring manager for this role: [paste ad]. Ask me the 8 questions you'd actually ask a junior [track] candidate, one at a time. After each of my answers, tell me what was strong and what a real interviewer would still be unsure about. Be a tough but fair interviewer.
The outreach structurer (you write the hook)I want to reach out to [studio/person]. My genuine, specific reason for wanting THEM: [you write this — the part AI can't]. Structure a short, warm, non-generic outreach message around my hook. Keep my voice. No "I hope this email finds you well", no "I'm excited to", no flattery. Under 120 words.
07Lab: Generic Detector + Case Study Rewrite (70 min)
Cold start friendly. Bring a real project; no project? Use your strongest coursework.
- Generate + detect (15 min): generate the generic case study; run the Generic Detector; highlight every stock phrase. (Seeing AI write your generic version is the wake-up call.)
- Extraction interview (25 min): answer with real specifics and stakes; get the rewrite. This is the assessed deliverable — it must read like you.
- Deflate + detect again (15 min): challenge every claim; replace surviving stock phrases by hand. ✏️ the lines only you could have written.
- Prompt pack (15 min): customize at least three job-search prompts (role matcher + interview rehearsal + outreach structurer recommended).
Zero-AI track (fully equal)
Partner-run extraction interview (genuinely better interview practice), hand-written case study, peer red-pen for the Generic Detector.
08Integrity Check
This week's four
- Does your rewritten case study contain a sentence only you could have written? If not, keep extracting.
- Did you deflate every claim to what's true and defensible in an interview?
- Is your AI disclosure specific (what AI did, what you did) — not a vague "AI-assisted"?
- Your prompt pack uses AI for research/structure/admin and keeps the voice and judgment human.
09Troubleshooting
| Problem | Fix |
|---|---|
| The rewrite still sounds generic | The interview wasn't specific enough. Re-run demanding "the detail that would only be true of MY project" |
| I don't have impressive results | Good — don't invent them. Process and decisions are the story; "I'd do X differently" shows growth recruiters value |
| AI keeps adding "passionate" / "excited" | Ban them by name. They're the loudest AI-voice tells |
| The outreach message feels fake | You let AI write the hook. The hook is yours — the specific reason you want THEM |
| Interview rehearsal feels too easy | "Be a tough but fair interviewer who pushes back on vague answers" — and answer out loud |
| I'm anxious AI is taking these jobs | Name the real skill: directing and verifying AI, not "using" it. That's the rarer, hireable thing |
10Resources
- ArticleLinkedHow to Write Design Case Studies That Get You Hired — IxDFThe case-study-as-story structure this week applies — the path is the juicy part. Discipline-agnostic despite the UX framing.
- ArticleLinkedCase Studies for Designers — The FuturProfessional case-study craft from the client-services world; pairs with the extraction interview.
- ExamplesLinked10 Exceptional Portfolios with Breakdowns — DesignerUpReal annotated examples — study what specificity and process documentation actually look like.
- TemplateLinkedPortfolio AI Disclosure Language (this course)Copy-and-adapt disclosure starting points, calibrated by how much AI did.
- ArticleVerify before teachingRecruiter perspective on AI in applications (current)The "every cover letter sounds the same now" reality from the hiring side; pick a current credible take — this shifts fast.
- VideoVerify before teachingA current creative-careers-and-AI talk (credible educator)For the career-reality framing; honest "what's actually changing" videos date quickly — choose the current best.
Optional extension
- Rewrite a second case study for a different kind of project — the extraction questions change with the work.
- Run the role matcher on a real job ad you'd apply to; act on the "2 things to learn" honestly.
- Hard mode: write your whole-portfolio disclosure paragraph — one honest statement of how you use AI across your practice. Increasingly something studios ask.
Next week: Capstone Sprint → — build the personal AI workflow, tool, or system that ties the whole course together.