The Proof-Burden Inversion

Your rubric scores which roles are worth winning. It tells you nothing about which ones you can win this week with what's published at junochen.com right now. I swept the active target set for the gap between your current evidence inventory and each company's proof requirements. The distribution inverts your priority list.

Four patterns, ranked by how much they should change your outreach sequence.

Pattern 1: Three proof tiers, and they run against your excitement

Sort the active set by how much inferential work the buyer has to do when they look at your published portfolio. Three tiers fall out.

Direct evidence. The mandate maps to a published case study or live artifact without the buyer connecting dots. Headway's Staff Product Designer, Group Practices is a permissions-heavy multi-party provider system. Thermo Fisher's decision gates and Allē's dual-surface platform address that directly. Stripe's Senior Staff, Risk is enterprise trust at transaction scale. Alibaba's $50B GMV redesign and Carrier IQ's review-and-approval surfaces speak to that problem without explanation.

Analogical evidence. The mandate contains a dimension your published work addresses sideways. Capital One's Director, AI in Experience Design wants org-wide AI transformation across a hundreds-person design function. Your Trust essay and Agentic Labs demonstrate the methodology. Alibaba demonstrates enterprise-scale change. But you don't have an org-transformation case study, so the buyer has to build part of the bridge themselves. Maven's VP of Design adds brand leadership and healthcare product ownership — your regulated-platform proof covers a slice, but recent healthcare brand stewardship is a gap. Vanta and Amplitude sit here too.

Missing evidence. The mandate depends on forward-looking artifacts in your build queue that aren't published yet. OpenAI's Product Design Lead, Growth–Codex is about designing when users specify intent rather than actions — making AI's potential become durable everyday value for business users. The Intent-Based Interaction artifact would address this directly. It doesn't exist as a linkable page. Midjourney's Design Lead has the same dependency and the same absence.

Your highest-rubric-score targets cluster in the missing-evidence tier. Your strongest-evidence targets are the ones you'd describe as solid but not thrilling. Your instinct Monday morning will be to reach out to OpenAI first. The evidence says reach out to Headway and Stripe first.

Pattern 2: One artifact would move three targets out of the missing-evidence tier

None of the four forward-looking artifact domains — Inference-Aware UX, Intent-Based Interaction, Agent Infrastructure as UX, Human-Agent System Design — has a published, linkable artifact at junochen.com. The Trust essay is live. The Agentic Labs apps are reachable at their standalone URLs. But the visual design deliverables that would demonstrate frontier thinking to an OpenAI or Midjourney hiring manager don't exist as discoverable pages.

I noted in Issue #8 that Capital One and OpenAI might have enough existing evidence to move, with targeted artifacts becoming interview prep rather than application prerequisites. The evidence check doesn't support that. Until at least one artifact domain ships, every Tier 3 target stays in Tier 3.

You can't build all four this month. Ship Intent-Based Interaction first. It directly serves both Tier 3 targets — OpenAI's Codex mandate sits squarely in the space of designing when users specify intent rather than actions, and Midjourney's creative-AI interaction problem lives there too. It strengthens the Capital One analogical case by demonstrating the AI design methodology you'd install across their org. And it connects to the TinyFish cold-start proof you already have a verbal answer for — the blank-box problem is an intent-setting problem. One artifact, three targets graduated or strengthened.

Pattern 3: The portfolio surface undermines even the ready tier

Tier 1 targets have direct evidence available. The portfolio surface doesn't deliver it.

junochen.com's homepage leads with TinyFish under "Selected work." Brand Pulse, Retail Velocity, and Carrier IQ — the three Agentic Labs apps that serve as your strongest AI-native production proof — aren't linked from the homepage. They exist at standalone URLs, but a hiring manager who lands on junochen.com and scans the work cards won't find them.

Headway and Stripe both face a routing problem. The buyer clicks through, sees TinyFish first (which can't serve as portfolio evidence because it's not a published case study), and has to work to locate the evidence that would actually convert them.

Three conditions compound across the tiers: evidence that doesn't exist yet, evidence that exists but requires buyer inference, and evidence that exists but isn't routed to the buyer. The routing fix and the essay removal are weekend tasks. The artifact gap is weeks.

Pattern 4: The buyers are incumbents and the paths are cold

One design executive joined a target company in the last 90 days: Banu Guler became Midjourney's Chief Design Officer through the Co–Star acquisition in late July, roughly six weeks ago. Everyone else — Daniela Jorge at Capital One, Ian Silber at OpenAI, Jason Lee at Maven, Ellen Dong at Headway, Jeremy Epling at Vanta, Amy Thibodeau at Gusto — has been in seat considerably longer. Most are a year or more into their tenure. Gab Menachem joined Amplitude as CPO in April, the most recent after Guler but still outside the 90-day window.

This changes the message type. A newly installed leader is still building — they're receptive to "here's how I'd help you build." An established incumbent is executing against a plan they already own. For six of seven identified buyers, your message needs to demonstrate that you understand what they're already doing and can accelerate it. Different posture.

Mirror their published language back to them. Daniela Jorge's design org narrative at Capital One uses specific framing about AI adoption within a mature design practice. Ian Silber's podcast appearances use specific language about composable primitives and the blank-box problem. Lead with their words. Guler at Midjourney is the exception — she's early enough in seat that a "here's how I think about the problem you're inheriting" message still has a window. It's closing.

The contact-path distribution compounds this. No verified warm introductions exist to any of these buyers. Stripe's Risk-design hiring owner remains unresolved entirely. The set is predominantly cold LinkedIn or direct application. Your portfolio surface and cover letter carry the full weight of first impression. No warm introduction creates context before the buyer encounters your materials. Which makes the Pattern 3 routing fix a precondition: fix the surface before you send any cold outreach, because cold outreach without a functioning portfolio converts to nothing.

Stripe's unresolved buyer: apply through the posting and state in the cover letter that you'd welcome a conversation with whoever leads design for Risk. The application becomes the buyer-identification mechanism.

What to do Monday

Sequence by proof readiness, not rubric score.

This weekend: remove the password-gated essay. Surface Agentic Labs on the homepage. Move TinyFish out of the lead position. These are preconditions for any outreach.

Then contact Headway and Stripe first — your published evidence directly addresses their mandates and both roles are live. Capital One moves after the routing fix, because the analogical evidence gets stronger when the portfolio surface actually delivers it. OpenAI and Midjourney wait until Intent-Based Interaction ships, because contacting them before it exists means asking the buyer to imagine evidence you could show them instead of showing it.

Anthropic: the current design board shows a single Product Designer role on the Evals & Prompts team, not a design leadership opening. Monitor until a qualifying role reappears.

Once Intent-Based Interaction ships, run this sweep again. The sequence will change.