Scan window: July 25 – August 1, 2026. Shelf life: seven days on Act Now, fourteen on Act Soon.
Upgraded: HackerOne (Deprioritize → Act Now). A live application endpoint outranks a missing feed listing.
Downgraded: Amplitude and Vanta to Act Soon. Ramp to Deprioritize. Elapsed time did all three. No mandate weakened and no evidence turned against them; the public applicant queue simply filled in. Treat these as a change in channel, not a change in quality.
Expiring:
| Signal | Leaves hot band | Note |
|---|---|---|
| Capital One Accelerator (posted Jul 14) | Aug 4 | — |
| DocuSign CPO — Graham Sheldon, started Jul 6 | Aug 5 | Crosses 30 days in seat. Nothing qualifying has surfaced under him. Let it go unless something does. |
| Webflow CPO and SVP AI (announced Jul 7) | Aug 6 | Company-reported only. |
| Gusto, Capital One Card Servicing | Aug 12 | Both fall out of the hot 21-day posting band. |
Gusto sits third despite the stronger role score, because there is nobody there to write to.
Send HackerOne today. Message Maven before you apply there. Get the Gusto application in within 72 hours.
Direct scans that produced nothing qualifying: Anthropic, OpenAI, Brex, Rubrik, Adobe. Two exceptions, both dateless. Abridge has a live Staff Product Designer on clinical AI notes with linked, auditable evidence ($216,750–$255,000). Atlassian has a live Principal Product Designer, AI Mobile, on Rovo agents. Neither exposes an original post date, so neither gets a slot this week. Resolve the dates next cycle before you spend a message on either.
Act Now
1. HackerOne — Director, AI Product Design | status revision
What changed. The direct Ashby page loads and serves the application form as of this morning. The role does not appear on HackerOne's public Ashby feed, which is currently carrying engineering, legal, and sales. The LinkedIn listing has expired.
Correcting the record. I killed this one last issue at 49 days on the strength of that feed omission. Bad call. An absent listing is weak evidence, because companies routinely unlist a role to throttle inbound while a slate closes. A working form is strong evidence. The right status was verify. It runs about 56 days old now, and 43-plus is normally a hard deprioritize, so read this as an override with cause rather than a rule I quietly dropped. Archetype, since this one got filed two different ways inside a single issue: AI-native.
Fit: Act — company 11 / role 15 · Window: unresolved, treat as closing · Move value: the act of entering the search is also the act of resolving its phase
Why it matters. This is the densest decision-rights mandate on the board. Agent-to-agent interaction, human oversight, transparency, accountability, reporting into the CPO. Somebody there owns the moment a person accepts, overrides, or reverses what a machine did. That moment is what "Trust Is the New Interface" (Jun 2026) is about.
Confidence. High on the buyer and the mandate. Speculative on search phase. I found no recruiter post, no announcement, and no hire.
Move
- (a) Who: Nidhi Aggarwal, Chief Product Officer. In seat since April 1, 2025, roughly sixteen months. Her Dec 2025 post on hybrid versus agentic security makes the argument that machines supply scale, humans supply context and judgment, and neither one carries the job alone.
- (b) Warm path: None apparent. Cold.
- (c) Send:
Nidhi — your argument that the machines bring scale and the people bring judgment is the design problem I've spent the last two years inside. I published "Trust Is the New Interface" in June: a watch/verify/delegate ladder for deciding when a system is allowed to act on its own and when a person has to put their name on the outcome. The insurtech quote-automation product in my Agentic Labs work sits on that same seam, where the machine's output stops and someone's judgment has to pick it up. I'm Head of Product at TinyFish, an enterprise web agent platform, so agent traces and reversibility are what I argue about on a Tuesday. Is the Director, AI Product Design search still taking candidates, or has the slate closed?
- (d) Don't: Keep the bug bounty marketplace out of it. She was brought in to rebuild the portfolio around AI. Marketplace suggestions land as a candidate who researched the previous company.
Tier Playbook: AI-native, before you send.
2. Maven Clinic — Senior Staff Product Designer, Care Delivery [IC track] | scope unresolved
What changed. The Care Delivery role is live and the hiring manager is publicly recruiting for it. Separately, Maven's VP of Design seat is open; the departing leader stated in July 2026 that he wrote the job description for his own backfill.
Fit: Act — company 10 / role 14 · Window: two to three weeks, closing the day the VP is hired · Move value: one question separates a director-shaped IC seat from a capped one
Why it matters. The posting says the hire defines how AI communicates uncertainty and hands control back to human care, and that those confidence and handoff patterns belong in the design system. With the chair above it empty, nobody outside the building can tell whether Care Delivery authors those standards or inherits them. Ask the question and you get a second answer for free: whether the VP requisition is the one you should actually be chasing.
Confidence. High on the vacancy and the hiring manager. Moderate that the ambiguity is real rather than sloppy copy.
Move
- (a) Who: Loric Avanessian, Senior Manager, Product Design. At Maven since April 2022. Her Figma case study, Oct 2025 covers a clinic finder that sat in the backlog for two years until she put something clickable in front of leadership, after which it shipped in under four sprints. Her stated position: AI prototyping pushes designers deeper into taste and strategy instead of eroding judgment.
- (b) Warm path: Cold, though she is actively posting about this requisition, which lifts response probability well above a standard cold note.
- (c) Send:
Loric — the clinic finder story stuck with me. Two years in the backlog, then something clickable in the room, then an MVP in under four sprints. That's the failure mode I keep hitting with AI in care settings: handoff behavior stays a slide until somebody builds one you can actually use. I published "Trust Is the New Interface" in June on how a system signals its own uncertainty and returns control to a person, and most of my 0→1 work has been in regulated environments, including Thermo Fisher's pharma supply platform. Before I apply: does Care Delivery own Maven's confidence and handoff standards, or apply patterns set above the team?
- (d) Don't: Leave the open VP role out of a first cold message to her. Asking a Senior Manager about the executive seat directly above her tells her you see her as a routing step.
Tier Playbook: Healthcare/Regulated.
3. Gusto — Head of Design, Unified Service Platform | fresh posting, no named buyer
What changed. Posted July 22, 2026. Ten days old. The requirements name uncertainty, graceful failure, human escalation, and AI outputs that shouldn't be sent.
Fit: Act Soon by rubric, Act Now by window — company 9 / role 13 · Window: hot through August 12 · Move value: low. No question to resolve, only a queue to get ahead of
Why it matters. "Outputs that should not be sent" is your essay's thesis written into a job description by people who have never read it. The soft spot is company AI centrality: at Gusto, AI is a strategic bet rather than the product. Write a two-pillar letter. Open on Agentic Labs and the trust ladder. Close on Thermo Fisher, because this is a new platform seat and somebody on that panel is privately wondering whether you can stand a function up from nothing.
Confidence. High. Primary ATS record, unambiguous copy.
Move
- (a) Who: No reporting line in the posting, and no design leader for this platform surfaced across Greenhouse, LinkedIn, or Gusto's design publications this pass. Application-first is the correct instrument here, not a fallback.
- (b) Warm path: Cold.
- (c) Cover letter opener:
Your posting names the case I care most about: the AI output that shouldn't be sent at all. I published "Trust Is the New Interface" in June on exactly that decision, a watch/verify/delegate ladder for when a system acts, defers, or stops, and my published Agentic Labs work in retail field intelligence and insurance quote automation is built around those escalation surfaces.
- (d) Don't: Don't show up as a horizontal platform leader. The mandate is service-domain, and horizontal framing reads as angling for a bigger job than the one on the page.
Tier Playbook: Growth-stage platform.
Act Soon
Babylist — AI Builder cluster [IC track on Staff/Senior] | seat count unresolved
Fit: Watch-plus — company 8 / role 12 (Director) · Window: two to four weeks · Move value: high. Nobody outside the company knows how many seats are actually funded
Six URLs, three mandates, mirrored across the U.S. and Canada. Director owning Registry, leading three to four designers, reporting into a VP/Head of Product Design and Research. Then Staff and Senior, written in the plural, with no product area attached to either.
Watch the Staff requisition. It previously ran as Staff Product Designer, Registry at $196–240K. Same requisition ID, now a broad "AI Builder" mandate at $207,666–$249,166 with a 20% target bonus. Read the band as a scope tell, not a number to plan around. When a req keeps its ID and changes its product area, its copy, and its pay, the first search usually taught the committee something they didn't know when they wrote it. High confidence on the mutation. Moderate on that reading.
Move. The reporting-line holder isn't publicly identifiable, so route to the nearest named product partner: Emily Levada, VP, Consumer Product. Her Jun 2026 post on AI transformation lessons argues for deterministic code anywhere reliability matters, written after watching agents pull the wrong year and summarize a dataset off ten rows. That's your opening. Cold.
Emily — your line about keeping deterministic code wherever reliability actually matters, written after the agent pulled the wrong year, is the most useful thing I've read from a consumer product leader on agents this year. I led design and research for Alibaba.com's B2B platform in North America through a homepage, search, and PDP rebuild, $50B+ GMV and +20% daily transactions, so I've built the version where a commerce surface has to earn trust at volume. I'm Head of Product at TinyFish now, building enterprise web agents, which is where I learned how much of "AI reliability" is a design decision about what you display and what you refuse to generate. I'm looking at your AI Builder roles: is the Director seat the Registry mandate specifically, with Staff and Senior sitting on other surfaces?
Don't: Skip the category enthusiasm about baby registries. Her public thinking is all execution reliability, and warmth about the category reads as filler.
Tier Playbook: Growth-stage platform. Battlecard candidate.
Capital One — Product Design Director, Card Servicing | 10 days old
Fit: company 6 / role 11 · Window: hot through August 12 · Move value: low
It clears Watch on role score alone, carried by scope and craft depth across internal servicing platforms, authoring environments, and governance orchestration. The company side is weak and will stay weak. No AI centrality, no equity window. Application-first, no public buyer surfaced. Lead Thermo Fisher and Red Cross. The agentic pillar misreads this room.
Card servicing is governance work before it is interface work. I built Thermo Fisher's mySupply platform across six pharma partners to 100% adoption and consolidated six Red Cross disbursement systems into one for national deployment, both cases where the audit trail was the product.
Tier Playbook: Enterprise platform.
Amplitude and Vanta | mandates intact, buyers unidentified
Fit: Amplitude company 10 / role 13; Vanta company 10 / role 13 · Window: two weeks before both cross 42 days · Move value: moderate, capped by the missing name
| Company | Role | Posting age | Design org | Scope |
|---|---|---|---|---|
| Amplitude | Head of Product Design | 25 days | ~15 | AI-native |
| Vanta | Head of Design | 26 days | ~40 | AI-native |
Neither mandate weakened. The competition arrived, that's all. I could not surface a hiring manager for either one this pass across ATS copy, leadership pages, and LinkedIn, which makes buyer identification the first job next cycle. Until then, apply. Don't wait for the name.
Amplitude opener:
Your posting treats human–agent workflows as the next shape of the product, which is the subject of the essay I published in June: a watch/verify/delegate ladder for when a system acts, defers, or stops.
Vanta opener:
Compliance is the one domain where a model's confidence has to be legible to an auditor and not just to a user, which is the governance-visible workflow problem I built for at Thermo Fisher.
One caution. It's tempting to read the Amplitude posting as the design requisition that trails a new product executive by 30 to 90 days. The timing fits the pattern, but I could not source the executive or the date this pass. Speculative. Don't build a cover letter on it.
Tier Playbooks: Growth-stage platform (Amplitude), Enterprise platform (Vanta).
Watch
- Babylist's VP/Head of Product Design and Research. No public identity yet. Escalates to Act Soon the day a profile or appointment post surfaces, because the Director reports into it. Growth-stage platform.
- Maven's VP of Design. Escalates on announcement. A new design executive's first thirty days beat any live posting on their own board. Healthcare/Regulated.
Deprioritize
- Ramp, Director Product Design. 54 days, no repost. Remove. Come back only if it reposts with changed copy, which would tell you what the first search missed.
- Salesforce, VP Product Design, Informatica. 78 days. Remove, strong scope notwithstanding.
- Givebutter, Director of Product Design. Deadline was July 20 and the page still loads. A page that renders is not an open search. Remove.
- Stripe, Senior Staff Agentic Commerce. URL returns 404. The live Link role is a separate search and proves nothing about this one. Revisit in six weeks.
-
Referrals still convert fourteen times better: Ashby's analysis of 38 million applications found 40% of referrals reach interview versus 3% of inbound applicants, which is the strongest argument for treating an aged posting as a channel problem rather than a dead one.
-
Capital One's July design cluster: Beyond Card Servicing, the bank posted Director, Product Design, Accelerator on July 14 plus Concierge Products and a Senior Director of Content Design — three distinct mandates, which reads as portfolio expansion rather than one flexible seat.
-
Two CPOs without design postings yet: CrowdStrike named AJ Shipley Chief Product Officer on July 15 to lead security for agents and agentic workflows, and no qualifying design role has followed — watch the 30-to-90-day window before the field notices.
-
Brex is no longer a pre-IPO equity story: Capital One's 8-K records the completed acquisition at roughly $2.56 billion cash plus 10.6 million shares, which retires the stage-and-equity assumption that put Brex on your original target list.

