The Evaluation Frame
Maven Clinic, Amae Health, Ontra, Aurora Solar, Front. Five companies, one unstated screen. Most candidates fail it before the portfolio deck loads.
The screen: does this person treat domain constraints as the design material, or as friction between them and the work they actually want to do?
Clinical protocols, insurance eligibility logic, contract negotiation hierarchies, compliance audit trails, provider burden, pricing interdependencies. These are the wireframes. A candidate who frames regulated, operationally complex work as "designing within limitations" has already told the panel they see the domain as something to survive rather than something to design with.
I've watched this evaluation run dozens of times at companies in this maturity band. The panel rarely articulates the criterion out loud. They just feel it. One candidate talks about the eligibility-check logic as the thing that shaped the intake flow. Another candidate talks about working closely with legal to ensure compliance. Both sentences describe the same activity. Only one candidate advances.
Confidence on this pattern: high. Maven's Head of Design and Consumer Design Lead postings foreground eligibility, claims, documentation, escalation, and provider burden. Aurora Solar's Head of Design posting centers systems-level design problems across residential and commercial solar, the kind of product surface where pricing interdependencies, installer workflows, and operational complexity are inherent to the design material. Neither posting mentions "product vision" as a standalone competency. That absence is the loudest signal in the tier.
How This Tier Evaluates Differently
Growth-stage panels scan for taste, velocity, and product instinct. Enterprise-platform panels scan for systems thinking and scale narrative. This tier scans for something else entirely, and the difference shows up in the first three minutes of a portfolio review.
At a growth-stage company, the panel is reading your visual judgment and product instinct in the opening slides. Here, they're reading whether you can narrate the operational logic behind a flow. Why this hierarchy, why this data appears at this moment, why this exception surfaces here instead of there. If your first three slides are polished screens without operational narration, you've already lost the room.
Exception-first thinking. What happens when the claim is denied, the contract term is rejected, the eligibility check returns ambiguous, the provider enters data that conflicts with the audit trail. If your case study shows the clean flow and mentions exceptions as an afterthought, you've told them you don't understand their product.
Multi-role awareness. These products serve users with fundamentally different needs looking at the same data. Provider and member. Caseworker and finance officer. Fund manager and legal reviewer. The panel wants evidence you designed one system serving multiple roles through shared data and role-specific views.
Regulatory visibility. The panel wants to see that compliance requirements shaped your information architecture. "I worked with legal to ensure compliance" tells them someone else handled the hard part. "The two regulatory regimes determined the dashboard hierarchy" tells them you operated inside the constraint and made it navigable.
Org Reality
Reporting lines matter more here than in most tiers. Maven's Head of Design reports to the CPO. This is the structural signal that determines whether the role is leadership or elevated execution. Design-to-CPO means design has a seat at the product table but not an independent one. Your authority depends entirely on how the CPO treats the relationship.
Ask this in the first conversation at Maven: "What decisions does the Head of Design own that the CPO does not override?" The answer, and the speed with which it arrives, tells you whether this is a strategic-partner relationship or a delivery-function relationship wearing a leadership title. Do not wait until the second round to ask. By then you've already invested positioning energy in a role whose mandate you haven't verified.
Design maturity across this tier: early-to-mid. These are companies where engineering and clinical or domain operations were built first. Design is being professionalized now. That's the mandate: hiring, process, design-system foundations, cross-functional credibility earned through shipped work. If you want to inherit a functioning team and focus on craft elevation, this tier is wrong.
Amae Health is the starkest case. Series B ($25M in November 2025, total funding over $50M), roughly 31 open roles, zero design positions posted. Clinical, patient operations, growth, and technology dominate the hiring board. Design as a named function is not yet visible. If you approach Amae, you are pitching yourself as the person who builds design from zero inside a company whose identity is clinical delivery and a precision medicine platform built on Palantir Foundry. That's a specific mandate. Want it specifically or skip it.
Ontra shows a similar gap. No design leadership role posted. Their Product Operations Manager posting references Design as one of four teams (alongside Product, Engineering, and Data) being operationally systematized. Design exists but its authority and reporting line are not publicly legible. Ontra is trusted by more than 1,000 global clients with contract automation for private markets. The constraint vocabulary here is contract negotiation logic, compliance hierarchies, and human-in-the-loop AI for legal work.
Front deviates. More below.
Pillar Mapping
Every recommendation maps to published work on junochen.com. No exceptions. Never reference TinyFish as portfolio evidence.
Maven Clinic
Lead with Allē. Support with Red Cross.
Confidence on this mapping: high. Maven is a dual-surface company with posted roles that spell out the operational vocabulary. Members and providers. The Allē case is the cleanest structural match: 30M members, 40K practices, one information model with role-specific views. Frame it operationally: "I designed systems where the same loyalty logic runs both surfaces, and the provider sees context — points, tier, expiry, birthday — before the appointment, not after." That sentence tells Maven you understand provider-burden reduction, which is the operational problem their postings foreground.
Support with Red Cross for eligibility-check and audit-trail logic. Three intake fields triggering eligibility checks, identity verification, and FEMA audit trail generation automatically. Same design problem as Maven's eligibility and claims workflows, different regulatory wrapper.
Amae Health
Lead with Red Cross. Support with Thermo Fisher.
Confidence on this mapping: moderate. The structural match is strong, but Amae has no posted design role and no visible design mandate. This is prospecting intelligence, not interview prep. If Amae posts a design role, re-evaluate the mapping against whatever operational vocabulary their posting reveals.
Amae's care model is multi-provider, multi-discipline (psychiatrists, therapists, dietitians, social-services specialists) serving patients with severe mental illness across multiple clinic locations. Their AI-enabled clinical features (Patient-at-a-Glance, Point of Care Assistant, Symptomology Phenotyping) suggest the design problem is coordination across roles under clinical protocols, with AI surfacing status changes and treatment recommendations that providers approve or adjust. The Red Cross case maps directly: one record for three roles that had not previously shared a system, deployed under operational pressure, designed for an untrained user on day one. Swap "field volunteer" for "peer support specialist" and the structural problem is identical.
Support with Thermo Fisher for regulated-operations credibility. Every design decision sat downstream of two regulatory regimes, and the platform made those regimes visible rather than hiding them.
Ontra
Lead with Thermo Fisher.
Confidence on this mapping: moderate. Ontra has no posted design leadership role, so this is prospecting positioning. The structural match between Thermo Fisher's exception-first operational design and Ontra's contract-automation workflows is strong.
Ontra's product is contract automation for private markets: AI-powered negotiation, playbooks with preferred/fallback/final positions, human-in-the-loop legal review with more than 600 legal professionals and rules-based quality checks. The Thermo Fisher case leads because it shows exception-first design in a regulated operational environment. Partners learning about exceptions at the delivery gate, where cost was 3-5x higher than catching the issue early. Translate directly: Ontra's users are catching contract issues late in negotiation cycles. Your instinct is to surface the exception before it becomes expensive. The batch-timeline and exception-visibility work in the mySupply case demonstrates exactly this.
Aurora Solar
Lead with Thermo Fisher. Support with Allē.
Confidence on this mapping: moderate-high. Aurora has a posted Head of Design role, which gives more signal than Amae or Ontra. But the posting language is broader than Maven's. Aurora's posting foregrounds systems-level design problems and cross-functional partnership without the granular operational vocabulary (eligibility, claims, documentation, escalation) that Maven's postings provide. Broader language gives you more room to frame the conversation but less certainty about what the panel foregrounds.
Thermo Fisher leads for operational complexity and exception handling. The mySupply dashboard work across OTD, Right First Time, OTIF, forecast accuracy, and capacity heatmaps maps to the kind of data-rich operational surfaces Aurora's solar-installation workflow requires. Pricing and payment complexity in solar involves installer margins, financing options, utility rate structures, and incentive calculations. Your Thermo Fisher narrative about making regulatory regimes visible rather than hiding them translates: Aurora's pricing interdependencies are the design material, the thing your portfolio should narrate, not the business context you summarize before showing screens.
Support with Allē because Aurora serves both solar installers (providers) and homeowners (consumers). Dual-surface company, one information model, two role-specific views. Frame the Allē case around the shared information architecture.
Front
Lead with Red Cross. Support with Thermo Fisher.
Front's Director, Product Design posting asks for comfort in queues, routing and assignment, rules builders, automation, and analytics. The Red Cross case shows triage logic (urgency bars replacing caseworker judgment under 10x surge) and multi-role consolidation in a workflow-heavy system. That maps to Front's routing and assignment problems directly.
Front is horizontal, which changes how you frame the constraint-literacy argument. Position it as transferable methodology: "I've designed triage and routing logic in systems where the rules changed under operational pressure and the users couldn't be retrained. That's the same problem Front's customers face when they configure queues and assignment rules across industries." One sentence. It connects your regulated-environment work to Front's horizontal flexibility requirement without pretending you're a vertical specialist.
Support with Thermo Fisher for operational-dashboard and data-rich workflow credibility. Front's posting also asks the leader to evolve and govern the design system and operationalize an AI-first design process. These are relevant but secondary topics. Never the opener.
Confidence on the Front mapping: high. The posting language is specific enough to match against, and the Red Cross case hits the operational vocabulary directly.
Anything emphasizing consumer product taste, visual design craft, or brand-led design thinking. The primary drift risk is Allē: if your framing slides toward "beautiful consumer app redesign" or leads with the 3.2x redemption number as a growth metric rather than as evidence that operational logic drove behavior change, you've lost the thread. Keep every Allē reference anchored to the shared information model, the provider-context workflow, and the operational mechanics behind the conversion numbers.
What Gets You Killed
"I love working within constraints." Sounds humble. Fails because it positions constraints as external impositions you gracefully tolerate. The evaluator hears: this person will need to be taught our domain before they can design in it. Fix: talk about constraints as the design input. "The eligibility-check logic shaped the intake flow" carries a fundamentally different signal than "I worked within the eligibility requirements." The first tells the panel the domain ran through your design decisions. The second tells them you accommodated it.
"I'd love to bring a fresh perspective to healthcare/legal/energy." Fresh perspective is a growth-stage value. This tier wants domain fluency, or at minimum, demonstrated ability to absorb domain complexity fast and design with it. If you lack direct healthcare or legal experience, frame your Thermo Fisher and Red Cross work as proof you've operated in regulated, operationally complex environments and the domain transfer is structural, not aspirational.
Leading with design-system work. Front's posting mentions design-system governance explicitly. Tempting opener. Don't. Design systems are table stakes at this level. Lead with workflow-design judgment, mention the system as infrastructure that enabled velocity. The exception: if you're past the first screen at Front and the conversation has moved to operational specifics, design-system governance becomes relevant. Never the opener.
"I want to build a world-class design org." Every candidate says this. It communicates nothing to a hiring panel at a company where design is still earning its seat. Replace it with something the panel can verify against your track record: "I've built design credibility inside organizations where design was not yet trusted, and I did it by shipping work that solved operational problems the business already cared about." The Red Cross case is the proof. A platform that handled 1,689 cases in two weeks, replaced six systems, and earned credibility through operational impact.
Pattern Breakers
Front is the clear deviation. A horizontal B2B platform serving Uber Freight, Navan, Stripe, whose posting reads like a vertical one (workflow-heavy, data-rich, queues, routing, rules, automation), which is why it's in this tier. But Front's design leader will need to design for horizontal flexibility across many industries rather than deep constraint literacy in one vertical. Recognize this in conversation. If your interviewer asks about how you'd approach a feature that serves both a logistics company and a fintech company, you're in horizontal territory. Lead with methodology.
Amae Health partially breaks the pattern. A healthcare delivery company whose primary product is in-person psychiatric care across clinic locations. Technology (the Palantir Foundry-based precision medicine platform, the AI clinical features) supports clinical delivery rather than replacing it. If Amae posts a design role, read the language carefully for physical-space signals versus software signals.
Ontra breaks on domain vocabulary. Private-markets legal operations is far enough from healthcare that the surface-level constraint language changes entirely. Contract negotiation playbooks, fund-lifecycle compliance, entity management. The structural evaluation frame holds. Do not use healthcare-adjacent framing in Ontra conversations.
Two-Directional Assessment
You need to evaluate this tier as hard as it evaluates you.
What this tier offers. Function-building mandates with genuinely hard design problems. The chance to professionalize design inside companies where operational complexity is real and the work matters. If you want to build a team, establish process, and earn cross-functional credibility by shipping work that solves problems the business already cares about, this tier delivers that.
What this tier risks. Structural marginalization. Design reporting to CPO (Maven) or embedded inside Product/Engineering without clear authority (Ontra, Amae) means your mandate depends entirely on a relationship. If that person leaves or priorities shift, your authority can evaporate. I've watched this happen repeatedly at companies in this maturity band. Eighteen months building credibility, the CPO departs, the new CPO has a different model, and you're rebuilding from zero with less patience than you had the first time.
Red flags to monitor:
- Any company in this tier that cannot articulate what their design team shipped in the last six months. Vague answers mean design is not yet operational enough to lead.
- Amae and Ontra have no posted design leadership roles. Approaching them is prospecting, not responding. You are educating them on why they need this function, which is a fundamentally different conversation than proving you're the right person for a defined role. The timeline is longer and the power dynamic is different.
- Front's posting has been live since May 15, updated June 22 (confirmed from Ashby page metadata). A posting that survives five weeks and gets revised means the first candidate pool didn't convert. The revision is the hiring committee's learning document. Whatever changed between versions is what they discovered they actually need. Confidence on this inference: high for the timeline, moderate for the implication. Postings also get revised for administrative reasons. But five weeks without a fill at this level, followed by a language update, is more often a recalibration than a copy edit.
Decision-makers at all five companies are likely offline Thursday through the weekend. Send outreach today or early Wednesday. Anything landing Thursday morning sits in an inbox that won't get real attention until Monday, buried under everything that accumulated over the Fourth of July break. Use the downtime to read Ontra's contract automation page and Amae's care model language closely enough to speak their workflow vocabulary, specifically.
Priority ranking across the tier:
- Maven. Posted role, clear evaluation frame, your Allē and Red Cross cases map cleanly. Highest-confidence positioning match in the tier.
- Front. Posted role, your Red Cross case maps to their routing and triage problems, and the revised posting suggests they're recalibrating. The $245K–$295K salary range is posted, which reduces one negotiation unknown.
- Aurora. Posted role, strong Thermo Fisher match, but broader posting language means less certainty about what the panel foregrounds. Verify the design mandate's scope before investing deep positioning energy.
- Ontra. No posted design role. Worth tracking for when one appears. The structural match to Thermo Fisher is strong enough that you could move quickly if a posting surfaces.
- Amae. No posted design role, and the company's identity is clinical delivery first. The longest path to a design conversation, but potentially the most interesting mandate if it materializes. Monitor the Greenhouse board for movement.
The constraint-literacy frame gives you a genuine advantage across this tier because most candidates from consumer or platform backgrounds will not think to lead with it. The risk is mandate depth. Verify before you commit.
- Maven's AI-native expectations: Maven's Head of Design posting describes the role as "AI-native" and expects the leader to modernize design-org work from research to production using AI, which means your constraint-literacy lead needs an AI-workflow layer ready for the second conversation.
- Headway as adjacent signal: Headway's Provider Experience Design Director posting describes a single provider session touching scheduling, telehealth, documentation, measurement, insurance verification, and billing, making it the closest structural comparison to Maven's evaluation frame in the current market.
- Front's design-system governance scope: Front's posting asks the Director to evolve the design system across tokens, components, governance, and collaboration model to ship 2-3x faster with confidence, which could become a primary conversation thread if you advance past the first screen.
- Amae's clinical AI features: Amae's Point of Care Assistant monitors patient status, flags meaningful changes, and offers treatment recommendations while providers approve or adjust, a human-in-the-loop pattern that maps directly to your trust-under-consequence framing if a design role surfaces.