Phase 1 — The Moment
Stripe is assembling a Data Products team from scratch — a Product Lead, a Frontend Engineer, and this Senior Staff designer — around a problem the company didn't have eighteen months ago. Stripe's transaction data now serves two audiences with incompatible trust requirements: human finance teams querying dashboards, and agents querying APIs and acting autonomously. Eeke de Milliano, Head of Product for Data & AI, framed the mandate publicly: ownership of Sigma, Data Pipeline, and "how AI agents will interact with business data in real time."
The real mandate: Design the surfaces where Stripe's financial data becomes a decision — whether a merchant is reading a report or an agent is executing a workflow. Scope spans Data Products, Atlas, App Marketplace, Business Network, and Ecosystems and Startups. That breadth tells you this is systems strategy across Stripe's data layer, not a single-product IC seat.
Three signals in five months. Agent wallets launched in April with per-transaction human approval. The OpenRouter acquisition in August added multi-model routing. The $159B pre-IPO valuation means equity at Senior Staff still carries real upside. This role doesn't own Link wallets or model routing directly, but the Data Products team designs what merchants and agents see when they look at Stripe's data — the layer where trust in agent-mediated financial decisions gets established.
Posting age: 44 days (published August 14, updated September 25). Stripe's stated 100-day window runs through November 22 — 56 days remaining — but they reserve the right to close earlier. The September 25 update means the search is actively managed. Three roles posted for the same team on overlapping timelines is an org build-out, not a backfill.
Someone looked at this requisition two days ago. The team is being assembled. The designer hired third has less influence over product direction than the designer hired first.
Phase 2 — Portfolio Mapping
Intelligence layer check: passes. Stripe's Data Products exist to surface financial intelligence for human and agent decision-making. The Product Lead posting confirms the team "turns Stripe transaction data into understanding and action" and expects AI-native workflows including agentic data access and automated reporting. The Trust essay's five handoffs are the through-line. Entry point: the moment between what Stripe's data surfaces and what a merchant or agent decides to do with it.
Lead with the Trust essay and CarrierIQ.
CarrierIQ's approval gate lets a user inspect agent-produced context — structured intake, visible execution stages, evidence, result comparison — before authorizing a consequential financial commitment (a policy bind). Link's spend-request flow does something architecturally similar: merchant name, purchase context, amount, line items, and payment method surfaced before credential release. Both systems separate what the agent found from what the human authorizes.
They diverge after approval. CarrierIQ authorizes a bind; Link releases a payment credential for a later checkout attempt. The approval event and the merchant's acceptance are distinct states, and the user's interface needs to represent that distance honestly. CarrierIQ's design direction — structured intake, visible execution stages, evidence review, approval gate — extends naturally into treating every consequential status change as a claim carrying evidence, owner, timestamp, and explicit unknowns. That's how this role would need to think about the gap between "approved" and "settled" in Stripe's agent commerce flow.
Link's CLI documentation already exposes approval-policy rules that can authorize spend requests without manual approval, while the consumer page still labels those controls "coming soon." The system's actual capability has outrun what the user-facing surface can explain — a gap between developer affordance and consumer legibility. CarrierIQ was built to close exactly this kind of drift, making the system's real state visible and controllable at the user layer regardless of what the infrastructure already supports underneath.
Support with Alibaba. $50B+ GMV, +20% daily transactions, +2.2pt NPS. The mechanism that matters here: diagnosing a structural gap in a platform's trust architecture, building the mandate, executing across cross-functional sprints. Stripe's Data Products team is building trust infrastructure at comparable scale.
TinyFish framing. Most recently as Head of Product at TinyFish, she shipped AI-native enterprise tools in production — agent traces, auditability, attribution, governance in enterprise deployments. That production experience is what makes the Trust essay's handoffs credible rather than speculative. Bridge narrative: moved to product to build AI-natively from zero, returned to design to apply that depth to a specific high-stakes domain.
Role geometry note. The posting describes a strategic IC — systems thinking, zero-to-one, cross-functional influence, movement between product direction and interaction craft. No management, no hiring, no direct reports. Katie Dill's recent talk reinforces this: designers at Stripe shape product direction, prototype rapidly, invent new patterns. Position as strategic IC with craft depth. Subordinate team-building narrative entirely.
Hardest objection: no fintech or payments domain experience. At Senior Staff, domain fluency is a reasonable expectation, and Stripe's financial infrastructure is genuinely complex. But the role's core problem is trust architecture for agent-mediated decisions over financial data. CarrierIQ demonstrates that design discipline — structured intake, visible execution, evidence review, approval gate, post-approval state tracking — applied to a different high-stakes financial domain. Alibaba demonstrates it at platform scale. Domain-specific financial knowledge is learnable; the judgment about how to architect trust across agent-mediated systems is what's scarce in this candidate pool.
Contact routing. Katie Dill first. At Senior Staff level, the design-side decision-maker carries veto authority on craft and systems judgment, and she has a specific recent public signal to reference (her September talk on encoding intent). Eeke de Milliano second. She's the product-side buyer assembling the cross-functional team and has direct context on the design seat's priority. But reaching the product leader before the design leader risks positioning you as a product-org candidate rather than a design-org hire.
Phase 3 — The Outreach Package
First Contact Message — to Katie Dill, Head of Design (178 words)
Your September talk on encoding intent — the gap between what AI generates fast and what qualifies as finished — is the tension I've been working through in agentic financial systems, specifically the moment between what an agent surfaces and what a human is entitled to trust enough to authorize.
I'm a design leader with 13 years across enterprise trust platforms (Alibaba.com, $50B+ GMV) and high-stakes 0→1 builds at BCG Digital Ventures (Thermo Fisher, Red Cross, Allē). Most recently I shipped AI-native enterprise tools as Head of Product at TinyFish, working daily with agent traces, auditability, and governance in production. I returned to design because I want to apply that technical depth to a specific, high-stakes domain — and Stripe's Data & AI challenge is the one I keep coming back to.
I published a framework for trust in agentic systems and built three production apps that implement it. The problem your Data Products team is solving — agents interacting with business data in real time, merchants needing to trust what they see — is the problem I've been designing for.
Would 20 minutes make sense to explore fit?
Resume Framing Note
Lead the summary with strategic IC language: "Senior design leader specializing in trust architecture for AI-mediated decision systems at enterprise scale." Surface Alibaba metrics first ($50B+ GMV, +20% transactions, +2.2pt NPS) — this is the scale proof Stripe evaluators will anchor on. Follow with the 0→1 portfolio under BCG Digital Ventures attribution (Thermo Fisher, Red Cross, Allē) to demonstrate range. Position TinyFish as: "Most recently Head of Product at TinyFish (enterprise web agent platform, $47M Series A), shipping AI-native tools in production." Do not use team-building or hiring language — this role has zero management scope. Avoid "design-led transformation" framing; Stripe's design culture is already strong. Use "systems strategy" and "interaction craft" — both appear in the posting.
Cover Letter Hook
Stripe's Data Products team is designing for two audiences that require different trust contracts — merchants who read dashboards and agents who act on APIs — and the trust architecture that bridges them doesn't exist yet. I've been building that bridge: a framework for what an interface is entitled to present as verified, traced, or authorized based on what the system has actually done, implemented across three production apps and grounded in shipping enterprise AI agents at TinyFish.
Phase 4 — Window Summary
| Action | Deadline | What degrades without it |
|---|---|---|
| Send first contact to Katie Dill | September 30 | The September 25 posting update means the search is active this week; every day without contact is a day another Senior Staff candidate advances in the pipeline |
| If no response, contact Eeke de Milliano (Head of Product, Data & AI) as product-side buyer | October 7 | Cross-functional team assembly is underway; the product leader is hiring her counterparts and has context on the design seat's priority |
| Final application window | November 22 | Stripe's stated 100-day close, but they can pull it earlier; treat October 15 as the practical ceiling for first meaningful contact |
- Link's approval policy gap: The Link CLI documentation already exposes approval-policy rules that authorize spend requests without manual approval, while the consumer page still labels those controls "coming soon" — worth referencing in conversation as a concrete trust-architecture problem the Data Products designer would inherit.
- Dill on encoding intent: Katie Dill's September 25 talk described Stripe's progression from an MCP that read design documentation to a design-system-backed CLI with complete templates and flows, framing the designer's job as encoding intent and refusing to equate fast output with finished quality.
- De Milliano's product build-out: Eeke de Milliano's hiring post names Sigma, Data Pipeline, and real-time agent interaction with business data as the Product Lead's scope — confirming the design seat sits inside an active cross-functional build rather than joining a stable team.
- Stripe's agent commerce vocabulary: The Frontend Engineer posting describes high-performance charting, streaming telemetry, interactive visualization, and reusable data-visualization standards built in partnership with design — useful for understanding the technical environment the designer would work within.