You have five counter-narrative dossiers. Title. Consulting. AI credibility. Speed. Hybrid. Each filed under a feature of your background that triggers a question.
The evidence in those dossiers is sound. The filing system is wrong.
A hiring panel of three people can ask you the same seven words and mean three entirely different things. The CTO who says "Have you shipped AI products?" wants to know whether you understand model behavior at a technical level. The VP Product asking the identical question wants to know whether you can move at startup speed without enterprise ceremony. The CEO wants to know whether you're overqualified and will leave in eighteen months.
Same surface. Three fears. Three evidence selections. Answer the surface and the fear stays intact underneath it.
Select evidence by anxiety, not by question.
The routing layer
Every objection your background triggers maps to one of four buyer fears. Diagnose which one is active before you select evidence or choose phrasing.
The diagnosis lives in the follow-ups, not the opening question. Listen for clustering. Three process questions in a row: speed anxiety. A hypothetical that compresses time ("if we needed to ship in six weeks"): speed anxiety. A request for specificity about model behavior: technical depth anxiety. Organizational language ("how do you hire," "what's your approach to design ops"): function-building anxiety. Tone is the tell for seniority fit. Deference means they fear you're overqualified. Probing means they suspect you're misleveled.
What follows is the diagnostic method for each archetype, with evidence mapped to your published portfolio at junochen.com, confidence-labeled, and compression-tested.
One calibration standard applies to every recommended phrasing below: would a confident senior operator say this out loud, in a room, to a peer? If the sentence sounds like it was written for a LinkedIn post or a coaching worksheet, it fails. If it sounds like something you'd actually say to a CPO across a table, it passes. Apply that test to anything you generate beyond what's here.
1. Speed anxiety
What the buyer fears: You will import enterprise alignment rituals, sign-off chains, and twelve-month timelines into a company that ships in weeks.
The grain of truth: You have led design at Alibaba and Thermo Fisher. Those are large organizations with real process overhead. A buyer who has been burned by an enterprise hire who couldn't adapt is not being irrational. They are pattern-matching against a failure mode they've seen before.
Diagnostic tells: Follow-ups cluster around process and timelines. "What does your first 90 days look like?" "How do you typically run a design sprint?" Hypothetical restatements that compress time: "So if we needed to ship in six weeks…" They are pre-screening for drag.
Evidence selection: Lead with Red Cross and Thermo Fisher. Red Cross was built under crisis-response constraints where delay had human cost. Thermo Fisher delivered five modules across six partners, nine sites, and three continents in twelve months with a team of one designer and two engineers. That is scope compression under real constraint, not enterprise bloat. Equinox+ adds a different speed register: venture-building pace at BCG Digital Ventures, concept to market without an existing org to slow it down.
Support with the three canonical Agentic Labs cases. Brand Pulse, Retail Velocity, and Carrier IQ are working prototypes, not decks. The prototypes themselves are the speed evidence. Mention TinyFish verbally as current-role context for AI shipping velocity. Do not point to it as portfolio proof.
Recommended phrasing:
"The builds I'm proudest of had the highest consequences for getting it wrong and the shortest timelines. Thermo Fisher was twelve months, three continents, one designer. Red Cross was crisis infrastructure. Equinox+ was venture-built from zero. I don't import process. I import judgment about which process to skip."
Confidence: high. Grounded in published case evidence with specific timelines and team sizes. A recruiter compresses this to "fast in high-stakes environments." That compression is accurate.
2. Technical depth anxiety
What the buyer fears: You are applying an AI label to traditional design work. You have not actually built products where agent behavior, model output, or trust calibration were core design problems.
The grain of truth: Your AI work is recent. The Agentic Labs cases are prototypes, not production systems at scale. A buyer at a company where models are core infrastructure has a legitimate question about whether your AI experience is deep or decorative.
Diagnostic tells: Questions probe for specificity. "How do you think about trust in agent-generated output?" "What's your experience with prompt design?" "How do you handle model failures in the UX?" The buyer is listening for whether you describe AI work from the inside or from a conference talk. Second tell: they ask about your Agentic Labs work and follow up with "And in a production environment?" That follow-up is the anxiety surfacing.
Evidence selection: Lead with the three canonical Agentic Labs cases and the trust essay. Brand Pulse, Retail Velocity, and Carrier IQ demonstrate hands-on design of agent-driven interfaces where the core design problem was calibrating human oversight of autonomous system behavior. The trust essay articulates a design philosophy for human-AI supervision that is specific enough to be falsifiable.
Support with the Thermo Fisher agentic vision. The case page closes with a reframe where five supply-chain modules become five continuously running agents, while batch QA release remains a human gate because a regulatory signature is irreplaceable. That vision demonstrates where you draw the line between agent autonomy and human authority inside a domain where the stakes are real.
Recommended phrasing:
"I haven't shipped from inside a frontier lab. What I've built is the trust layer. Carrier IQ is real-time risk scoring where the human makes the call. Thermo Fisher's agentic vision keeps a human gate on QA release because a regulatory signature can't be delegated to a model. The design problem I solve is: when should the system act, when should it surface, and when should it stop."
Confidence: high for the Agentic Labs and trust-essay evidence. Use with caution if the buyer's definition of "technical depth" means ML engineering or model training. You are not an ML engineer. Do not let the reframe drift into that territory. If you hear questions about training data, fine-tuning methodology, or inference optimization, acknowledge the boundary cleanly. Do not blur it.
3. Function-building anxiety
What the buyer fears: You are a strong individual contributor or consultant who has never built, hired for, or scaled a design organization. You can do the work but cannot build the function.
The grain of truth: Your most recent role is Head of Product at a small company. The consulting chapter is visible. A buyer scanning quickly could read the trajectory as IC → consultant → product-titled role and never find the org-building.
Diagnostic tells: "Have you built a design team from scratch?" "How do you think about hiring designers?" "What's your approach to design ops?" Questions about culture, retention, career laddering. The tell is organizational language. They are not asking what you designed. They are asking what you built around the designing.
Evidence selection: Lead with Alibaba and Allē. Alibaba was Head of Design and Research, North America, for U.S. and European buyers at $50B GMV. That is a function-leadership title with geographic scope. Allē was Product Design Director leading three product designers plus UX research and brand across two surfaces (consumer iOS and provider web) serving 30M+ members and 40K+ provider practices. Two surfaces, two user types, a team with multiple disciplines.
Support with Thermo Fisher. Product Design Director coordinating across six partners and nine sites. The coordination complexity is the evidence, even though the direct team was small. Use Thermo Fisher for coordination scope, not team scale.
Recommended phrasing:
"At Alibaba I led design and research for North America. At Allē I ran a team across consumer and provider surfaces for 30 million members. At Thermo Fisher I coordinated design across six pharmaceutical partners and nine manufacturing sites. I've built at different scales, and I know the difference between hiring your first designer and hiring your tenth."
Confidence: high for Alibaba and Allē. Moderate for Thermo Fisher as function-building evidence. A recruiter compresses this to "led design teams at Alibaba and Allē, coordinated cross-partner design at Thermo Fisher." That compression is clean because the evidence is concrete.
4. Seniority-fit anxiety
What the buyer fears: Either you're overqualified (will take the role and leave when something bigger appears) or misleveled (your title says Head of Product, not Head of Design, so maybe you're not a design leader at this altitude).
This one splits in two directions. The diagnostic tells are opposite.
Overqualified tells: Deferential tone. "What excites you about a company at our stage?" "We're a small team — would that be a step back for you?" They like your background. They are afraid you won't stay.
Misleveled tells: Probing tone. "Tell me about a time you influenced executive decisions." "How do you think about design's seat at the table?" They are testing whether your seniority is real or inflated by title.
The grain of truth for both: The overqualified fear is reasonable. You have led design at Alibaba. A 40-person startup should wonder why you want their role. The misleveled fear is also reasonable. Head of Product at a small company is ambiguous. Does it mean you expanded into product leadership, or does it mean design didn't have enough scope to justify a standalone title?
Evidence selection for overqualified: No case study fixes this. The fix is specificity about why this company's problem is the one you want. Generic enthusiasm is the worst possible answer because it confirms the fear. The difference:
Generic: "I'm really excited about what you're building in the agent space." This is what the recruiter writes down as "she said she's excited about the problem." It compresses to nothing. It gives the hiring manager no reason to believe you'll stay.
Specific: "Your exception-handling flow is the same design problem I solved at Thermo Fisher, except your version runs on live model output instead of manufacturing data. I've spent two years thinking about where the human gate belongs in that loop. That's the problem I want."
The specific version names the company's problem, maps your evidence to it, and explains why this role closes a gap you care about. It compresses to "she wants this specific problem because of Thermo Fisher." That compression holds.
Evidence selection for misleveled: Lead with the full portfolio breadth. Five published design leadership mandates: Alibaba, Thermo Fisher, Red Cross, Equinox+, Allē. The Head of Product title at TinyFish is scope expansion, not demotion. Use it as evidence that you operate across product and design.
Recommended phrasing for misleveled:
"The title is Head of Product. The work is design leadership with commercial accountability layered on. Five published mandates, each with shipped systems adopted in production environments."
Confidence: moderate. The overqualified reframe depends entirely on your ability to name something specific about the target company. If you can't, the reframe collapses. The misleveled reframe is stable but requires the portfolio to be clean.
Which is the problem that compounds all four.
Portfolio artifacts as anxiety amplifiers
Every archetype above assumes the buyer encounters your evidence cleanly. They don't.
Your portfolio currently forces the buyer to reconcile your story before they can evaluate your candidacy. That reconciliation labor compounds whichever anxiety is already active. A buyer who is already wondering whether you're fast enough hits a metadata inconsistency and now has two things to resolve instead of one. The anxiety doesn't wait while they sort out your site.
Do not make the buyer reconcile your story. Every minute they spend resolving inconsistencies is a minute spent inside whichever anxiety archetype is already active. Reconciliation confirms the fear rather than resolving it.
Here is what is creating reconciliation burden right now:
The browser-title metadata reads "Product Designer." This is the first thing a search partner sees when they open your site in a tab. If speed anxiety is active, it confirms you're an IC, not a leader who owns pace. If seniority-fit anxiety is active, it raises the misleveled question before you've said a word. The tab title is doing damage in a space you cannot see and cannot narrate around.
A TinyFish case card is visible on the site. The entire method above instructs you to use TinyFish as verbal current-role context and technical AI currency, never as portfolio proof. A visible case card contradicts that strategy. If function-building anxiety is active, a buyer sees a case study for a small startup and wonders whether this is your highest-altitude work. If seniority-fit anxiety is active, the card makes the Head of Product title more prominent, not less. The card pulls TinyFish from the verbal layer, where you control the framing, into the portfolio layer, where the buyer draws their own conclusions.
Five Agentic Labs entries appear on the homepage, but only three resolve cleanly. Brand Pulse, Retail Velocity, and Carrier IQ are your canonical set. Creator Scout returns a 404. UAT Sentinel redirects to a login gate. If technical depth anxiety is active, a buyer who clicks through your labs and hits a 404 and a login wall on two of five will conclude the AI work is thinner than presented. Absence is neutral. Breakage is negative signal. Two broken links do more damage than two missing links would.
The field-note essay is still exposed. "What Do You Count When the Playbook Doesn't Exist?" returns HTTP 200 with a client-side overlay that says "Private Essay." The full essay text is in the HTML source. It names TinyFish's CEO and CPO, describes pricing decisions, references agent-step pricing, and discusses billing infrastructure migration. A thorough evaluator who views source will find it. As of today, July 19, this exposure is still active. The client-side gate is not a gate. It is a suggestion. If technical depth anxiety is active, finding startup pricing mechanics in viewable source reframes your AI work as small-company commercial operations, not frontier design problems. If seniority-fit anxiety is active, finding confidential client details exposed behind a cosmetic gate raises a judgment question that has nothing to do with design skill and everything to do with whether you handle sensitive information carefully. Neither reading helps you.
Fix the browser title. Remove the TinyFish case card or reclassify it as verbal-only context. Remove or fix the two non-canonical lab entries. Move the essay behind a real server-side gate or take it offline. These are positioning infrastructure, not polish tasks.
The reference-safety standard
A reframe that works in your mouth but breaks in someone else's summary is a liability.
At Director+ level, your story passes through at least four compression points before a hiring decision is made. Each one strips nuance. The question is whether what survives compression still resolves the anxiety, or whether the nuance was the part doing the work.
Compression point 1: Recruiter paraphrase. The internal recruiter hears your answer and retells it to the hiring manager in one or two sentences. Your entire reframe becomes "she said she's fast" or "she said she builds trust layers."
Compression point 2: Written search-partner summary. A16z's executive hiring process states that search firms circulate written reports before candidates meet the team. Practitioner sources describe these as one page per candidate with ranked fit reasons, gaps to probe, and recommended interview questions for each gap. Your career compresses into three bullet points and a flag. The flag is the anxiety archetype. The bullet points are whatever survived compression point 1.
Compression point 3: Live interview. You control this one. But you control it inside a frame already set by the first two compression points. If the written summary flagged "consulting background, probe for long-term commitment," the interviewer walks in with that anchor loaded.
Compression point 4: Backchannel reference check. Backdoor references are increasing as employers respond to AI-polished applications. Someone not on your reference list gets a call. They describe you in their words, not yours. If your reframe uses language your former colleagues would never use to describe you, the backchannel surfaces the gap.
How each reframe performs under compression:
- Speed reframe compresses to "fast in high-stakes environments." Survives. A former colleague would recognize "judgment about which process to skip" as true.
- Technical depth reframe compresses to "designs human-AI trust interfaces, not ML infrastructure." Survives, with a caveat: "trust layer" sounds abstract in someone else's mouth. Anchor it to Carrier IQ or Thermo Fisher's human gate every time you use it.
- Function-building reframe compresses to "led design teams at Alibaba and Allē, coordinated cross-partner design at Thermo Fisher." Survives cleanly because it is concrete.
- Overqualified reframe is the most fragile. It compresses well only if the gap you named was specific enough to be memorable. "I'm excited about the problem" becomes exactly that in the recruiter's notes, and it means nothing. "Your exception-handling flow is the Thermo Fisher problem on live model output" becomes that in the recruiter's notes, and it holds.
- Misleveled reframe compresses to "five design leadership mandates, current role adds commercial scope." Stable.
The 8:30 AM routing card
You're reading this before you walk into a room. Here is the compression:
Speed → Follow-ups cluster on process and timelines → Lead with Red Cross and Thermo Fisher (twelve months, one designer, three continents) → "I import judgment about which process to skip."
Technical depth → Follow-ups probe model behavior and production specifics → Lead with Carrier IQ and the trust essay, support with Thermo Fisher's human gate → "I build the trust layer. When should the system act, when should it surface, when should it stop."
Function-building → Follow-ups use organizational language (hiring, ops, culture) → Lead with Alibaba (Head of Design, North America) and Allē (three designers, two surfaces, 30M members) → "I've built at different scales."
Seniority fit → Deference means overqualified, probing means misleveled → Overqualified: name the specific problem, not generic enthusiasm → Misleveled: "Five published mandates. The title is Head of Product. The work is design leadership with commercial accountability."
Diagnose first. That's where most candidates lose the room.
-
AI-native roles expect craft proximity: Ramp's Director of Product Design posting asks leaders to ship PRs, design memory, and rethink process as new capabilities unlock, which means the hybrid objection is becoming less of an objection and more of a baseline expectation.
-
Gusto's escalation design questions: Gusto's June 2026 AI principles post names three open design problems including showing what an agent is doing, designing intelligent escalation, and preserving human agency when automation makes consequential decisions, all of which map directly to the technical depth reframe above.
-
Trust explanations can backfire: Microsoft Research found that feature-based AI explanations did not improve human decision-making and actually increased overreliance, which means "I make AI trustworthy" needs mechanism language, not explanation language.
-
DoorDash validates the dual evaluation: DoorDash Design's practitioner guidance confirms that design-leader portfolio presentations now evaluate both team operations and product innovation, so function-building and craft evidence need to appear in the same conversation, not separate ones.