The strong background is the baseline. What separates the candidate who advances from the one who doesn't is whether the second paragraph resolves the specific anxiety sitting in the buyer's head before they finish reading it.
Same candidate. Four positions. Four letters. For the buyer's purposes, four different people. Each framework below opens with the evidence that kills the dominant doubt in that room, uses only proof traceable to junochen.com, and is structured around the decision the buyer is actually making. In "Same Anxiety, Different Room" I mapped how identical portfolio evidence lands differently depending on who evaluates it. This turns that mapping into paragraphs you send.
Diagnose your tier from the posting language. Grab the hook, proof block, and closing ask. Swap bracketed references for the actual target. Match the posting's vocabulary at the sentence level. Submit.
Tier 1: Frontier AI / Intelligence Layer
Archetype companies: OpenAI, Anthropic, Amplitude, Glean, analytics and agent platforms.
The anxiety you're solving: Her AI work is decorative. She designs interfaces about AI, not the trust architecture inside agentic workflows. She has never built in the medium.
How to spot it. The verb "built" appears more than "designed." The posting names agents, models, or orchestration rather than generic product design scope. If it asks what a human needs to see, decide, or override when a system acts on their behalf, you are in Tier 1.
Lead with the essay, not a case study. The essay IS the credential.
Opening hook:
I published "Trust Is the New Interface" in June because I kept watching the same failure pattern across agentic products: a system returns a confident result, the user has no basis to evaluate it, and the product team calls the ensuing distrust a "UX polish" problem. It is an architecture problem. [COMPANY]'s work on [SPECIFIC PRODUCT CHALLENGE — e.g., "Wave's agent-surfaced opportunities" / "tool-use orchestration across multi-step tasks"] sits at exactly the point where this architecture either earns trust or loses it, where a user decides whether to act on what an agent produced.
Proof block (2–3 sentences, named systems, capability metrics):
The essay names five agentic workflow handoffs — intent-setting, in-progress visibility, output review, decision gate, and loop feedback — and I have built working systems against each of them. Carrier IQ runs five insurance carrier agents simultaneously and returns live quotes with source evidence and side-by-side comparison in ninety seconds; the design problem was giving an operator enough provenance to verify a result without slowing the loop. Brand Pulse runs parallel agents across Reddit and X, classifies mentions by sentiment and theme, and builds a live brand score as signals arrive; the design problem was showing work as it builds rather than leaving the user at a blank wait state. At TinyFish, where I currently lead design across three AI products shipped in three months, I see these handoff failures daily, which is why the labs and the essay come from operating context, not theory.
Closing ask:
Where in [PRODUCT/WORKFLOW] is the human-agent handoff the binding design constraint right now — the place where users have the information to act but not the confidence? That is the specific problem I have been building against, and I would like to hear how your team frames it.
Trust Proof Element:
My portfolio practices what the essay argues: the Trust essay is public and unguarded because the argument should be accessible; the full case studies use progressive disclosure from teaser to full evidence, because the work requires context to evaluate well. The architecture mirrors the trust ladder the essay describes — watch, then verify, then go deeper.
⚠️ Activate after boundary fixes ship. Tier 1 buyers are the most technically literate audience in the landscape. If a hiring manager or engineer inspects the site and finds client-side-only gating on case studies, the progressive-disclosure claim becomes a liability instead of a proof point. Hold until server-side access control is live.
Tier 2: Regulated Platform
Archetype companies: Headway, Gusto, Ramp, Vanta, health/fintech/compliance platforms.
The anxiety you're solving: She hasn't operated where errors carry real consequences. Scale is one problem. A mistake that affects someone's healthcare access, payroll, or compliance status is a categorically different one.
How to spot it. Compliance, regulatory, patient/member/provider, risk. The posting names specific user types on both sides of a platform. If "trust" appears in the context of sensitive data or regulated workflows rather than AI agents, you are in Tier 2.
Lead with Red Cross or Thermo Fisher mission-critical proof. Allē for multi-sided buyers.
Opening hook:
The hardest design problems I have solved were hard because the cost of getting it wrong was measured in someone's safety or access to care. At the American Red Cross, I redesigned the family-registration workflow that caseworkers use during active disasters, where every field error delays a family's access to shelter and aid resources. At Thermo Fisher, I designed the quality-release gate where a batch decision affects patient safety across nine pharmaceutical manufacturing sites — and the system had to work identically at every one of them. [COMPANY]'s [SPECIFIC PRODUCT/USER CONTEXT — e.g., "provider-matching workflow" / "payroll compliance surface"] carries the same weight: the user on the other side of the screen is depending on the system to get it right.
Proof block (2–3 sentences, named evidence, structural proof):
At Thermo Fisher, the decision gate I designed — where five agents run continuously but a human makes the release call with full audit trail and confidence context — became the model for how the organization handles high-stakes automation across its global site network. For Allē, I designed the dual-surface loyalty architecture serving both consumers tracking rewards and providers managing treatment records, two user types with fundamentally different trust requirements sharing one platform, using a decision-block structure that makes each surface's logic legible without collapsing the distinction. At TinyFish, I currently lead design for AI products where the stakes are commercial rather than clinical, but the pattern is the same: the system has to earn the right to act before the user will let it.
Closing ask:
What is the decision your users are most afraid to get wrong inside [PRODUCT] — and how much of that fear comes from the product versus the regulatory context around it? That distinction shapes everything about how I would approach the trust architecture.
Trust Proof Element:
My portfolio is structured the way I believe regulated platforms should be: public-facing information is clearly scoped, while sensitive case evidence sits behind progressive disclosure from teaser to full case, matched to the viewer's need-to-know.
⚠️ Activate after boundary fixes ship. Lower risk than Tier 1. Regulated-platform buyers are less likely to inspect site code. But "structured access" as a claim still depends on the gating being real. Hold or soften to the language above until server-side controls are live.
Tier 3: Enterprise
Archetype companies: Salesforce, Atlassian, Adobe, large-platform design organizations.
The anxiety you're solving: She cannot survive the matrix. Enterprise design leadership is a political-survival problem as much as a craft problem. Candidates from smaller companies consistently underestimate how much of the job is cross-functional negotiation inside organizational constraints they did not choose and cannot change.
How to spot it. The role spans multiple organizations. "Cross-functional collaboration" appears more than twice. Scope is defined by platform surface or business unit, not the whole company. If the posting names more than three partner functions, you are in Tier 3. Current example: Adobe's GenStudio Director of Product Design asks for design vision around AI-first agentic workflows and measuring adoption, value, and trust. The pattern recurs across Salesforce, Atlassian, and similar orgs where the design title masks a matrix-navigation mandate.
Lead with Alibaba trust architecture at scale.
Opening hook:
At Alibaba.com, I led the trust-architecture redesign for a B2B marketplace serving millions of cross-border transactions where the buyer and seller have never met, speak different languages, and operate under different regulatory frameworks. The design problem was building enough visible trust infrastructure — verified supplier signals, transaction protection, dispute resolution transparency — that a buyer in São Paulo would commit $50,000 to a manufacturer in Shenzhen based on what the platform showed them. [COMPANY]'s [SPECIFIC PLATFORM CHALLENGE] operates in the same territory: the design work that matters most is the trust architecture underneath the product surface.
Proof block (2–3 sentences, scope and organizational proof):
The Alibaba redesign shipped across the full buyer journey — discovery, evaluation, transaction, and post-purchase — at $50B+ GMV scale, driving +20% daily transactions and −47% buyer-reported security concerns, and I operated across product, engineering, data science, legal, and localization teams, starting with 32 stakeholder interviews to build the mandate before touching a screen. The design decisions that mattered most were the ones I negotiated into the roadmap, not the ones I made in Figma. I have done this at speed too: Equinox+ went from zero to MVP in ninety days across five brands, 600K+ premium members, a 4.8-star App Store outcome, and a data-driven IA where members who followed at least one instructor within thirty days retained at nearly double the rate of those who did not.
Closing ask:
Where in [SPECIFIC PRODUCT AREA] does design currently have the least leverage over decisions that affect the user — and is that a resourcing problem, a reporting-structure problem, or a trust problem with the cross-functional partners? The answer changes the first six months entirely.
Trust Proof Element:
My portfolio architecture reflects how I think about enterprise trust: the public essay establishes the framework, the case teasers show scope and outcomes, and the full cases are available through progressive disclosure from teaser to full evidence because enterprise-grade evidence requires context to evaluate responsibly.
⚠️ Activate after boundary fixes ship. Enterprise buyers are less likely to inspect site implementation than Tier 1 engineers, but "progressive disclosure" as a structural claim still depends on the gating being real. Moderate risk. Soften or hold.
Tier 4: Enterprise Design Org Transformation
Archetype: Capital One and similar organizations hiring a design leader to change how design operates, not to manage the existing function.
The anxiety you're solving: She can evaluate and critique, but can she build? Org-transformation roles attract candidates who are articulate about what is broken and have never actually rebuilt anything. The buyer needs proof of construction. Analysis without building is the thing they just fired.
How to spot it. Negation language. "Not an executive who manages up and delegates down." "Not design ops." "Not the approval committee." Builder language: "created (not inherited)," "stood up," "established." If the posting describes the operating model as the deliverable rather than the team as the deliverable, you are in Tier 4. The company is telling you what went wrong last time. Your letter mirrors that negation back as capability. (I covered why negation language is the real brief in Issue #4.)
Invert standard logic. Lead with Trust essay + live Agentic Labs as proof of building. Case studies appear as secondary org credibility only.
Opening hook:
I wrote "Trust Is the New Interface" to make an argument about products, but the argument applies to design organizations too: the question is whether the operating model is designed for the handoffs that AI creates. Intent-setting, in-progress visibility, output review, decision gates, loop feedback — these are not just product patterns. They are the patterns a design org needs when the tools, the workflows, and the definition of quality are all shifting at once. Then I built working systems against each of those patterns, live, with real APIs, in Agentic Labs.
Proof block (2–3 sentences, building proof, then org credibility):
Carrier IQ orchestrates five insurance carrier agents simultaneously and returns evidence-backed quotes with side-by-side comparison in ninety seconds. Retail Velocity discovers 247 independent SMB restaurants with one agent, scans live menus, and surfaces competitive displacement opportunities. These are live demos built with real APIs in the last three months at TinyFish, where I lead design across three AI products. The case studies behind them — Alibaba's cross-border trust architecture at $50B+ GMV scale, Thermo Fisher's pharmaceutical decision gates, Equinox+'s zero-to-MVP in ninety days for 600K+ members — prove I have done this at organizational scale. The labs prove I am doing it now.
Closing ask:
What does quality mean when the tools change every quarter? I want to talk about the operating model — how [COMPANY] wants design to work with product and engineering when the workflows themselves are being redesigned, not just the outputs.
Trust Proof Element:
The portfolio itself is the proof of building: a public essay that frames the problem, live agent systems that demonstrate solutions, and case studies structured with progressive disclosure from teaser to full evidence. I did not write about agentic design and then show traditional case studies. I built the systems the essay describes.
⚠️ Activate after boundary fixes ship. Highest risk tier. The entire framework claims proof of building. If the site's own access architecture uses client-side-only gating — a trust boundary any engineer on the hiring panel could bypass in thirty seconds — the meta-positioning collapses. Hold this element until server-side controls are live, or remove the access-architecture claim entirely and let the labs carry the building proof alone.
On the Trust Proof Element
The idea is the strongest meta-positioning move in the portfolio: a site that practices the trust philosophy it argues. Progressive disclosure from public essay to teaser to gated case study mirrors the trust ladder the essay names. The portfolio becomes its own proof.
The problem is the implementation. In "Your Site Is Making Arguments You Didn't Choose to Make", I walked through the current client-side access logic: it returns full HTML before the redirect fires. A portfolio arguing for decision gates while shipping a decision gate that does not actually gate is contradicting its own thesis in front of the audience most equipped to notice.
When the boundary fixes ship, activate the Trust Proof Element in all four tiers. Until then, the labs and the essay carry the building proof. They are strong enough. Do not make a claim the site cannot currently defend.
Agentic Labs metrics: what they prove and what they don't
The labs cite capability descriptors. Five carriers in ninety seconds. 247 venues with one agent. These are live demos built with real APIs across enterprise verticals, not shipped products with user bases. The capability descriptors prove you can build in the medium. The case studies prove you can ship at scale. The combination is the argument. Do not apologize for the absence of business metrics on systems designed to demonstrate architectural thinking. That is not what they are for.
- Capital One's negation language: The Director, Product Design - AI in Experience Design posting names Claude, Claude Code, Stately, XState, and Windsurf as relevant tools, making it the clearest Tier 4 target where the operating-model-as-deliverable framing maps directly.
- Gusto's AI-native design org shift: Gusto published a piece on how its design team became an AI-native org within a quarter, including MCP-connected design-system workflows, which is the closest public proof that a Tier 2 company is evaluating candidates on process transformation, not just regulated-product experience.
- Adobe GenStudio's source conflict: Adobe Design's careers page still shows the GenStudio Director of Product Design role as active with "Apply now," while the main Adobe careers page for the same requisition says the job has been filled — verify before using as a Tier 3 target.
- Vanta's 40-person org scope: The Head of Design posting spans GRC, Trust, Platform, Self-serve, and AI teams at $365K–$513K cash with equity, making it a rare role that straddles Tier 2 regulated-platform anxiety and Tier 4 org-transformation scope.