Six Verdicts Move Once the Whole Tracker Runs on One Arithmetic

Last week this section ran two scorecards for the same companies in one issue. Moves said Active Targets wins any disagreement between them, then carried forward October 16 numbers that Active Targets had already revised. Earlier issues had their own errors. They put Watch labels on totals below the Watch line, and they scored AI exposure at zero on a scale that starts at 1.

A label that doesn't follow from its own components can't tell you where to spend a week of search effort. So this issue adds no companies. Every carried-over entry that still scores Act or Watch has been recomputed under one set of rules. This week's new postings are in Moves, and Ramp's financing status is in Window Status.

The rules every entry is scored under

Scores. Company is out of 12, in this order: AI centrality (0–3), stage/equity, design influence, trajectory. Role is out of 15, in this order: comp, scope, craft depth, AI exposure, portfolio value. The numbers in parentheses after each score follow those orders. Act requires company 9+ and role 12+. Watch requires company 7+ or role 10+.

1. Baseline is the October 22 Active Targets scorecard, or the latest earlier print if that tracker didn't carry the entry. A component moves only on new employer evidence or on one of these rules. With no new data, the prior score stands.
2. The verdict follows the sum. Posting age runs on a separate clock that sets priority and never sets the label. "Watch for this cycle" is how a 10/14 ended up at Watch. A posting is hot for 21 days. Day 42 ends its priority window.
3. Comp scores base cash at the midpoint of the posted range for SF, NYC or US remote: 3 at $250K+, 2 at $200–249K, 1 below that. The top of a range describes the exceptional hire, so it doesn't count. Where an employer states where most hires land, that figure replaces the midpoint. Equity is scored once, by issuer, under stage/equity. † marks a prior comp score kept because no range was pulled.
4. AI exposure scores the product the role designs. It reaches 3 only at AI-native companies, and at AI centrality 2 it caps at 2. A posting that mentions AI only as a design-team tool doesn't lower it. An explicit agent mandate earns its credit under portfolio value.
5. Ownership events count once, under stage/equity. A pending take-private caps equity. It doesn't also drag trajectory.
6. Location is pass/fail and sits outside the score. US remote or Bay Area passes. A failing role keeps its visible score, but recommended effort drops to zero, so a high total can't hide a dealbreaker.

Stage/equity uses the issuer rule from issue #17:

• 3: private Series C–E at a fresh valuation.
• 2: private above $50B with employee tenders, a stale 2021–22 price, or public under two years.
• 1: public more than two years, equity capped by a pending deal, or seed stage.

Prior and recomputed verdicts, side by side

Three causes account for the six changes:

• OpenAI, Figma and Stripe Link already summed to Act. Their labels now match their numbers. Link wasn't on the October 22 tracker, so its baseline is the October 16 print, 11/12, labelled Watch. Its company score drops to 10 because the issuer rule moves Stripe's stage from 3 to 2. The role holds at 12. Scope falls to 2, matching the Checkout seat, and AI exposure rises from 1 to 2 for a team whose stated product direction is agent payments.
• Ramp moves on new evidence. The posting is now verified, and it gives design a real agent mandate.
• Stripe Checkout and Gusto move under the comp rule. Checkout's midpoint is $239K. Gusto misses the $200K line by $500, which makes it the most fragile call in the table.

Off the live tracker, nine Watch labels come off. Five drop to Ignore because their totals never reached a Watch line, and four have no matching posting. Axon/Prepared stays unresolved (see the long tail).

Where effort goes this week, in clock order

• By Tuesday, November 3: Figma and Anthropic Design Systems. Both are Act, and both hit day 42.
• By November 5: Runway, from the long tail: Act, live, US remote. Gusto's hot window closes the same day, and five openings put it first among Watch roles.
• November 6–12: Anthropic Safeguards (Nov 6), Developer Platform (Nov 9), Ramp (Nov 11), OpenAI Consumer Verticals (Nov 12).
• Stripe Link: Act at the threshold, on an older req. Use Dan Nelson's October 3 post as the route in. Checkout's hot window ends today.
• Watch, hot through November 2: Samsara first, then Talkiatry.
• No effort: Legora and Build fail the location gate. Hold EvolutionIQ and ServiceNow until their location terms are clear. Peregrine and Axon/Prepared wait for re-verification.

The thirteen live entries

1. Legora. Legal AI workspace that firms use for research, review and drafting (Series E).

• Stage / Capital: Series E, $550M at $8.8B post-money, October 21.
• Value chain: Downstream application on foundation models. Design owns the review surface a lawyer has to trust before signing off.
• Best at: Embedded product designers plus a central studio that carries work through to shipped code, a combination that is rare at this stage.
• Signals: The funding is nine days old and still hot. Every listed design role is onsite in NYC, London or Stockholm.
• Hiring manager: Anton Reimertz, Head of Design since August 2025, laid out the studio-plus-embedded model in an October 30 interview.
• Landscape context: Harvey is the obvious comparison. Our October 22 AI Capital piece read the Series E as a push toward in-house legal buyers, and that is where this seat's product will be tested.
• Rubric scores: Company 11/12 (3/3/2/3). Role 15/15 (3/3/3/3/3). Midpoint $278K.
• Verdict: Act, unchanged. Blocked by the gate. The posting's only explicit AI duty concerns the design team's own tools. AI exposure stays at 3 anyway, because every surface at Legora is the AI product. Revisit when a US-remote or Bay Area design role shows up in the feed.

2. Anthropic. Claude models, apps and developer platform (Series H).

• Stage / Capital: $65B Series H at $965B (May 28). The public S-1, filed October 26, discloses an April employee tender. It doesn't say whether new hires reach future ones.
• Value chain: Vertically integrated from model to app. Design sits unusually close to model behavior.
• Best at: Treating design guidance as something the model reads and gets tested against. The Design Systems role owns "the guidance, metadata and evals that show whether Claude uses it correctly."
• Signals: Four roles open: Design Systems, Safeguards, Developer Platform and Growth.
• Hiring manager: Joel Lewenstein, Head of Product Design since October 2023. No successor to Jenny Wen has been established. Hannah Hearth is publicly linked to Developer Platform design, though whether she owns that req is unconfirmed.
• Landscape context: Stage scores 2 whether or not the IPO happens, because private-with-tenders and recent-IPO fall in the same bucket.
• Rubric scores: Company 11/12 (3/2/3/3). Role 14/15 (3/2/3/3/3). Design Systems midpoint $345K.
• Verdict: Act, unchanged. Timing is the only constraint; work the roles in the clock order above.

3. OpenAI, Consumer Verticals. Designs ChatGPT experiences for health, finance and new agentic uses (posting).

• Stage / Capital: Private, well above $50B, with a history of employee tenders. Stage scores 2 under the issuer rule. Current valuation and IPO filing status weren't re-verified this pass.
• Value chain: Vertically integrated. This team designs the point where a general assistant enters regulated domains.
• Best at: The posting names "trust, accuracy, safety, and user agency" as design work, where most companies would leave them to policy.
• Signals: Published October 1. The Growth role is scored in Moves.
• Hiring manager: Ian Silber, Head of Product Design, talks publicly about hands-on prototyping and building from reusable components. The direct manager for this role is unconfirmed.
• Landscape context: A trust failure in health or finance costs more than one in open-ended chat, which is why the posting hands that work to design.
• Rubric scores: Company 10/12 (3/2/2/3). Role 14/15 (3†/2/3/3/3).
• Verdict: Watch → Act. The sum never supported Watch. If comp turns out to be a 2, the role still scores 13.

4. Figma, Director, Product Design, AI & Context. Leads the layer that feeds design context to agents: Design Systems, Code, MCP, CMS and Community (posting).

• Stage / Capital: Public (NYSE: FIG) since July 31, 2025, about 15 months.
• Value chain: Midstream tooling. This director owns what agents read, upstream of every AI feature Figma ships.
• Best at: The structured design data agents need to produce on-brand output.
• Signals: Updated October 27.
• Hiring manager: Loredana Crisan, Chief Design Officer since Q3 2025, stressed cutting tedium while keeping taste and play when she joined. Emily Campbell joined in October to lead AI Models design and works next to this role. She isn't established as its manager.
• Landscape context: The next earnings report will show whether AI features convert to paid usage, and this director owns the context layer those features depend on.
• Rubric scores: Company 10/12 (2/2/3/3). Role 13/15 (3/3/2/2/3). AI exposure drops from 3 to 2 under rule 4. Craft is 2 because a director seat sits further from the work. Midpoint $333K.
• Verdict: Watch → Act.

5. Ramp, Principal Product Designer, Design Systems Lead. Corporate cards, spend management and bill pay, increasingly run by agents (posting).

• Stage / Capital: $1.85B of new capital at $60B (October 15). An employee tender is documented from November 2025.
• Value chain: Downstream application that owns the finance team's spend workflow.
• Best at: Designing for finance teams and employees, two groups that distrust each other.
• Signals: Published September 30, now titled Principal. A Director role open since June 8 looks like a standing req, one kept open indefinitely for the right candidate.
• Hiring manager: Diego Zaks, VP Design and an early employee (author page), frames design as the mediator of that trust.
• Landscape context: With Brex inside Capital One, Ramp is the independent equity story in corporate spend.
• Rubric scores: Company 10/12 (2/2/3/3). Role 13/15 (3/2/3/2/3), midpoint $332.5K. Portfolio value rises to 3 because the role owns "how agents discover guidance, compose interfaces, and check their work," evals and code included.
• Verdict: Watch → Act. The old Watch was pending verification. The posting is now verified.

6. Stripe, Link and Checkout. Checkout is the hosted payment page. Link saves buyer payment details across merchants (Link, Checkout).

• Stage / Capital: Last new capital at $50B (2023). The February 2026 tender at $159B was open only to current and former employees.
• Value chain: Midstream payment rails. These two surfaces are where buyers see Stripe directly.
• Best at: Turning network scale into buyer-facing payment UX.
• Signals: Checkout was published October 9. Link is an older req that Dan Nelson promoted on October 3; earlier coverage misread that post as Nelson's own promotion. Data & AI now returns a 404, and Risk dates to July with no confirmed score. Both leave the tracker.
• Hiring manager: Katie Dill, Head of Design since March 2021. Yuliya Gorlovetsky is Head of Product Design (tenure unconfirmed). Nelson sponsors Link.
• Landscape context: Link's posting describes "AI agents transact on your behalf" as product direction, but it doesn't give design ownership of permissions or of when a human steps in.
• Rubric scores: Company 10/12 (2/2/3/3). Link 12/15 (3/2/3/2/2), midpoint $263K. Checkout 11/15 (2/2/3/2/2), midpoint $239K.
• Verdict: Link Watch → Act, exactly at threshold. Lose any one point and it returns to Watch. Checkout Act → Watch. The two roles differ only by $24K of midpoint.

7. Gusto, Staff Product Designer. Payroll, benefits and HR for small businesses (posting).

• Stage / Capital: The last confirmed valuation is the 2021 Series E at $9.5B. The June 2025 tender at $9.3B required two years of service.
• Value chain: Downstream application.
• Best at: Designers who ship code, a shift Amy Thibodeau drove.
• Signals: Five openings, published October 15. A Senior Staff, Finance role open since August 2025 is a standing req.
• Hiring manager: Amy Thibodeau, Chief Design Officer since 2021.
• Landscape context: Gusto is the only fintech entry that makes design responsible for how AI behaves when it's wrong: "representing uncertainty, surfacing errors gracefully, preserving customer agency, and knowing when a human should step in."
• Rubric scores: Company 9/12 (2/2/3/2). Role 11/15 (1/2/3/2/3). SF/NYC midpoint $199.5K.
• Verdict: Act → Watch. One comp point restores Act, and five openings make this the most forgiving req in the tracker. Work it first among Watch roles despite the label.

8. EvolutionIQ, Staff Product Designer, Agentic Platform. AI guidance for insurance claims, owned by CCC (posting).

• Stage / Capital: CCC's take-private by GTCR and Elliott is pending, expected in Q1 2027. Unvested CCC RSUs outstanding at close convert to restricted cash awards that generally keep their vesting terms. The posting offers RSUs without naming the issuer.
• Value chain: Vertical application inside CCC's insurer network.
• Best at: The strongest agent mandate on this list. Design owns "uncertainty, automation, control, intervention, and recovery."
• Signals: Q3 results (October 29): revenue up 10.1%, AI run-rate about $130M against $120M in Q2, guidance suspended.
• Hiring manager: Christy Wedge leads AI product design (start date unconfirmed). CPO Benjamin Berry is the fallback.
• Landscape context: The merger agreement requires the buyer's approval for hires above $300K base. A Staff hire likely sits under that line, so the deal probably won't freeze this req (moderate confidence).
• Rubric scores: Company 8/12 (3/1/2/2). Role 13/15 (2/2/3/3/3), midpoint $220K.
• Verdict: Watch, unchanged. On October 16 it read 9/12 and was labelled Watch when it should have been Act. October 22 capped stage at 1 for the take-private, but it also cut trajectory to 1 and raised portfolio to 3, giving 7/13. Rule 5 restores trajectory: 8/13, one company point short of Act. Gate unresolved: the posting lists New York and says nothing about remote.

9. Samsara, Staff Product Designer. Sensors, video and software for fleets and industrial sites (posting).

• Stage / Capital: Public (NYSE: IOT) since December 2021.
• Value chain: Vertically integrated hardware and software.
• Best at: Owning the physical-operations data its agents act on.
• Signals: Published October 12. Agentic features launched June 24.
• Hiring manager: Correction: the fallback contact is Johan Land, EVP and CPO. Kiren Sekar has retired. No named successor to former design head Ackermann has been established.
• Landscape context: Without a named design head, this req has the least visible leadership behind it of any entry. A named successor resets the entry.
• Rubric scores: Company 8/12 (2/1/2/3). Role 11/15 (2/2/3/2/2). Comp drops to 2 on a $200K midpoint across pay zones.
• Verdict: Watch, unchanged. The posting mentions AI only as the designer's own tool. AI exposure holds at 2 because the fleet product the role designs now ships agents.

10. ServiceNow, Sr. Staff Product Designer, AI/ML. Designs an AI product that prepares customer data for use by AI systems (ATS record).

• Stage / Capital: Public (NYSE: NOW) since 2012.
• Value chain: Midstream platform, with design scope across products.
• Best at: Distribution through IT and HR workflows that already run on it.
• Signals: Released October 8. The careers link returns a 404, but the ATS record is live. Q3 growth slowed to 20.5% from 24%, and the 10-Q shows $167M in severance year to date.
• Hiring manager: Dantley Davis, SVP Design (statement), pushes consumer-grade craft into enterprise software.
• Landscape context: Slowing growth plus severance is the pattern to watch at incumbents adding AI to existing platforms.
• Rubric scores: Company 7/12 (2/1/2/2). Role 12/15 (3/2/3/2/2), midpoint $262.5K.
• Verdict: Watch, unchanged. If trajectory fell to 1, the role score alone would still hold Watch. Gate unresolved: "Remote, WA (Seattle)" may mean Washington only.

11. Atlassian, Senior Director, Jira Design (24872). Design leadership for Jira, including its agents (posting).

• Stage / Capital: Public (Nasdaq: TEAM). A March 2026 restructuring cut about 10% of staff.
• Value chain: Midstream work-management platform.
• Best at: Owning the system of record where engineering work gets planned.
• Signals: The posting assigns "Jira's agent experiences and human-in-the-loop decisions."
• Hiring manager: David Hoang, VP Design for Rovo/AI since July 2024 (blog). Whether he owns this req is unconfirmed.
• Landscape context: Designing agent decisions inside a system of record is a stronger mandate than designing a standalone assistant.
• Rubric scores: Company 7/12 (2/1/2/2). Role 11/15 (2/3/2/2/2). Atlassian says most hires land near the base of the range, $221K, so that figure replaces the midpoint.
• Verdict: Watch, unchanged. Gate unresolved: confirm location eligibility before you spend effort here.

12. Talkiatry, Staff Product Designer. In-network virtual psychiatry (posting).

• Stage / Capital: Historically an LLC issuer, so expect equity as units rather than stock (unconfirmed).
• Value chain: Vertically integrated care delivery.
• Best at: Insurance-covered psychiatry at scale.
• Signals: Published October 12. The role defines how patients interact with conversational AI, including being open about its limits.
• Hiring manager: Matt Ivester, CPO (joined 2024; exact CPO appointment date unconfirmed), prioritizes patient access and clinician confidence.
• Landscape context: The clinical-answer companies compete over licensed reference content such as UpToDate. Talkiatry's AI sits inside a care relationship it already delivers.
• Rubric scores: Company 7/12 (1/2/2/2). Role 10/15 (1†/2/3/2/2). Stage holds at its prior 2 under rule 1, because no current financing price turned up to place it in an issuer bucket. At 3 or 1 the company score becomes 8 or 6, and role 10 holds Watch either way.
• Verdict: Watch, unchanged, low priority. Comp holds the score down. The gate passes (US remote).

13. Build, Design Lead. Company 8/12 (2/1/3/2). Role 11/15 (1/3/3/2/2). Watch on the numbers, but it fails the gate (onsite NYC) and is past day 42. No effort.

Older entries from October 11 and 16

These entries fell off the October 22 tracker without being retired, and several are still live. The arithmetic is corrected below, with rules 3 and 4 applied where data exists. Hiring managers weren't re-verified, so each entry returns to full format when its posting is.

Both of Peregrine's printed versions clear Act, so the version labelled Watch was the error. Axon/Prepared is harder. Its two versions disagree on AI centrality (3 versus 2) and on design influence, and no new evidence settles which is right. Rule 4 also takes a point of AI exposure off the lower version, leaving 7/12. One version reads Act and the other Watch, so the entry waits.

The five Ignores carried Watch labels their totals never earned. Each comes back on a new or re-leveled role. Mysten Labs exposes a quirk in the rubric: AI centrality works as a bonus, so a company with no AI in the product can reach Watch on stage, influence and trajectory alone. It is Watch on paper; give it effort last.