Trigger
Ramp's Director, Product Design posting hit day 31 unfilled while the company shipped five agent products in the last 90 days, including procurement agents, AP memory, and a full accounting-firm platform (confidence: high the role is live; moderate the extended window reflects a high bar, not a paused req).
Window
You're inside the 22–42 day urgency band. A high-quality probe now positions you whether the pipeline is still active or the posting reappears with revised language after day 42.
Either way, waiting costs more than moving. If the posting refreshes, you're already on their radar with the right frame.
What this trigger probably means for the design org
Ramp closed a $40B valuation round in May 2026. Since late April they've shipped AP agent memory, procurement AI agents, Ramp Stack, and Applied AI Solutions. Four major agent launches in seven weeks. The design org is under real load, and the posting's language confirms the direction: tools, agents, memory, PRs. Look at agents.ramp.com. Per-agent budgets, spend limits, exception approvals, expiring credentials, agent-specific cards for Claude and ChatGPT. Ramp is building financial identity and governance for non-human actors.
The design problem here is calibration. When an autonomous system moves money across 70K+ companies, what does the human need to see, when, and at what level of detail to trust the outcome without reviewing every transaction? That is a trust-architecture problem, and your essay maps it. Your labs are testing it in production.
Your angle
Lead with "Trust Is the New Interface" and Agentic Labs. The essay names the problem Ramp is hiring someone to solve: users anchor to worst outcome, not average accuracy. One confident wrong answer does more damage than ten correct ones repair. Carrier IQ (InsurTech quote automation, regulated domain) is the closest Agentic Labs parallel to Ramp's agent-mediated financial workflows.
Then Alibaba and Thermo Fisher as proof the capability survives scale and consequence. Alibaba ($50B+ GMV, +20% daily transactions) proves you can redesign trust signals across a platform where B2B procurement decisions carry real financial weight. Thermo Fisher matters here for the exception-first order management pattern: surfacing 3 orders that matter out of 197 is the exact design logic Ramp's agents need for spend exceptions. And the agentic rebuild vision (5 agents, one human gate) shows you've already architected the autonomous version.
Do not lead with speed-and-scrappiness narratives. Ramp needs someone who has thought structurally about what happens when the system acts autonomously and the human's role shifts from operator to auditor.
Who to contact
Diego Zaks, VP Design, Ramp. LinkedIn. One of Ramp's earliest employees. Prior: Director of Digital Design at Juul Labs. MFA, Pratt.
Public fingerprint: His Midday post (May 7, 2026) is your hook. He frames Ramp's design bar as "design to build memorable experiences, AI to handle the work on behalf of customers" and says Ramp "measures and celebrates less time spent on our product." He also wrote a 2021 piece on trust in finance teams, arguing well-designed automation should flag situations needing human review while handling the rest. Five years old, but the philosophy maps directly to your trust essay's framework. No recent conference talks or podcast appearances findable.
Warm intro path: Not established from public signal. Ramp's investors include Sequoia, Founders Fund, Thrive Capital, Coatue. Check your network for Ramp, Juul, Pratt, or investor-adjacent contacts. Default is cold.
Channel: Cold LinkedIn DM, anchored to the Midday post.
Outreach messages
Warm message (if you find a connection):
"[Name] suggested I reach out. I read Diego's piece on Midday joining Ramp, and his line about measuring success by less time spent in the product stopped me. That's the design problem I've been writing about. I published 'Trust Is the New Interface' in June on how agentic systems earn user trust through calibrated handoffs — the design layer that accuracy alone can't deliver — and I've been building live agent products (Carrier IQ, Brand Pulse) that test those principles in production. I'd love 20 minutes to hear how the design team is thinking about trust architecture as Ramp's agents take on more autonomous financial authority."
Cold message:
"Diego — your Midday post framed something I haven't seen anyone else in fintech design say clearly: the goal is less time in the product, not more engagement. That's the thesis behind my recent essay 'Trust Is the New Interface,' which maps how agentic systems earn trust through five calibrated handoff patterns. I've been building live agent products that test this (Carrier IQ automates insurance quoting in a regulated domain), and my Alibaba work ($50B+ GMV platform redesign) was fundamentally about making high-stakes B2B procurement decisions trustworthy at scale. I'd welcome a conversation about how Ramp's design org is approaching trust as agents take on spending authority."
What not to say:
- Do not frame yourself as a "fintech designer." Ramp's design mandate has moved past fintech into agent-infrastructure territory. The fintech label shrinks the conversation.
- Do not cite the 2021 trust post as your primary hook. It's useful context, but leading with a five-year-old blog post signals you couldn't find anything current. The Midday piece is the opener.
The non-obvious thing
Ramp's Browserbase customer story reveals how their procurement agent was actually designed. The product team watched a real customer perform a live vendor review step by step, then converted each observed manual action into an agent requirement. The first real test denied a software request within seconds after the agent independently visited the vendor site, pulled trust documentation, and found the tool wasn't SOC 2 compliant.
Two implications. First, Ramp's design process is workflow-observation-driven. Your research-to-architecture approach (32 warehouse floor interviews at Thermo Fisher, 32 buyer interviews at Alibaba) will land as native methodology. Second, the compliance-grounding of the agent's first real decision signals that trust and auditability are built into the design from the start. Your trust essay's framework maps directly onto what they're already building.
Warning signs
Zaks's background is advertising and branding, not product design. His LinkedIn still reads "Creative Director." Not disqualifying, but calibrate your vocabulary: lead with craft quality and user outcomes over systems-design jargon.
The posting's proximity to tools, agents, memory, and PRs raises a structural question worth probing: if design is being pulled this close to engineering, the role may carry less organizational authority than the title suggests. It could be a craft-execution seat inside an engineering-led product org rather than a design-leadership seat with real strategic influence. Ask early who else is on the interview loop and whether design has a seat in product-direction decisions or is downstream of them.
- Ramp's agent-economy surface: The agents.ramp.com microsite shows per-agent budgets, spend limits, exception approvals, and agent-specific cards, which is the most concrete public evidence of the trust-architecture design problem you'd be walking into.
- Headway's provider design mandate: Headway's Design Director, Provider Experience posting owns clinical documentation, AI-assisted notes, insurance verification, and billing across 75K+ providers, making it the strongest regulated-workflow parallel to Ramp if you want a second probe this week.
- AI Spend Intelligence signal: Ramp's April 2026 launch post reports 13× growth in average monthly AI token spend across customers since January 2025, which gives you a concrete data point for conversations about why financial governance of autonomous systems is accelerating.
- Ramp Stack's autonomy model: The Ramp Stack launch lets accounting firms decide where agents run autonomously versus pause for review, which maps directly to your trust essay's Watch→Verify→Delegate ladder and could be a specific reference point in conversation with Zaks.