VP of Product

The title says VP. The seat says founder: no team to inherit, no org to manage, one product loop to own end to end.

Product.ai is building the verified truth layer for shopping, the intelligence that tells you what's actually true about a product, including when not to buy. Our first proof at scale is SimplyCodes, the code verification service, at about $22M a year in revenue, profitable, bootstrapped since 2009. No outside investors. No board. Fewer than twenty operators outbuilding companies 10x our size.

Why This Role Exists

We are advertising one seat through several doors. One door reads Founding Product Lead. This one reads VP of Product, because some of the people who should take this seat have spent a decade earning that title and will not apply to a role that looks like a step down. The seat itself does not change: you own the consumer quality bar, the strategy-to-spec pipeline, and the verdict on what ships.

A VP title reliably pulls in people who expect a team to inherit, and that mismatch usually costs both sides a quarter before anyone admits it. There is no team here on day one. The founder holds company strategy and keeps it. You do not inherit an org chart or a headcount plan. You inherit a blank spec and a founder who will argue with you about what goes in it.

Your first hires, if any come, are earned by the product, not budgeted in advance. You prove the seat is worth headcount by shipping something that needs more hands, then you make the case for the hands. The pull here is scope: you decide what the company's next dollar of product comes from, with the economics of an owner rather than a manager.

The System You'll Need to Model

• Knowledge-graph products. One trustworthy answer combines what exists, what it costs, which claims survive verification with citations intact, and what a specific shopper cares about. Every product call is about how those layers combine, and they keep moving.
• Trust-calibrated conversational products. A general assistant averages the internet with confidence. This one is built to say a calibrated "don't buy this," and the product bar is whether a user trusts that verdict on a purchase that matters to them. Calibrated trust is harder than engagement, and engagement metrics will not tell you if you got it right.
• Agent-mediated distribution. Increasingly, an AI agent decides whether to call your product before any human sees it. Being chosen by an agent is becoming what being indexed by a search engine used to be, and you are writing that playbook while you ship against it. Verified commerce data already sells to developers and agents on metered keys, so the surface you own includes an API's product shape as well as a screen a person reads. The commercial relationship with those developers and platforms belongs to our commercial seat; the product shape of what they call belongs to you.
• Spec-driven agentic development. A product decision becomes a written spec precise enough for a coding agent to build from, and a separate agent grades the finished build against that spec rather than the builder's own word. The verifier is the crux: it is a check the building agent cannot write or pass for itself. Your spec is the interface the whole system builds and grades against, and your taste is the gate.
• A governed knowledge substrate. The company runs its own operating decisions through a shared AI brain that answers its own questions from thousands of internal documents, the same product family this seat sells externally. You will work inside the exact system you are building for customers, so every gap you feel is a gap a customer feels too.
• A self-amending rules engine. Since August 2026 any operator here can propose a change to how the company runs and ship it live, no founder approval required. You will need to model a governance system that keeps changing under you, the same way the product does, and your product decisions live inside that same current.

If reading that energizes you, keep going. If it feels overwhelming or underspecified, this isn't the right fit.

What You Will Own

• Product law. The locked specs coding agents build from, spanning conversational answers, the pages that carry those answers into search, and personalization. Agents write the code and the content; you write the spec, the acceptance tests, and the verdict on what ships. This is a daily production discipline here, not a pilot.
• The consumer quality bar. The standard for everything a shopper touches, written as falsifiable evidence tests for four to six outcomes, each with a check a stranger could run. When the bar and the schedule conflict, you hold the bar.
• The decision-shaped interface standard. What a verdict looks like on screen: an interface built around the decision a shopper is making, buy, wait, or walk away, not around whatever content happens to exist. Communicating a calibrated "we are sure" versus "we are not" is the hardest interaction problem in the product, and you own how it reads.
• One or two falsifiable outcomes a quarter, run by you personally. From intent to locked spec to verified build, with your name on the result either way. This is how you are measured: decisions registered and outcomes moved, not headcount grown.
• Your seat charter. Within your first quarter you co-sign a charter for this seat. It names one machine-checkable number that proves the seat is working, and a written split of what you decide freely versus what you bring to the founder.

You will use the craft you already own (product strategy and cross-functional influence with no reporting line to lean on) and grow into directing coding agents as your production system, verification design for outputs no human can grade at the machine's speed, and distribution built for AI agents as the customer instead of a search engine.

Who You Are

How you think. You form your own working model of a complex system, and when the locked spec you wrote turns out wrong, you catch it and rewrite it before anyone tells you to. You do not need a fully scoped brief to start; you need enough signal to reason from first principles and enough discipline to write down what you conclude. Clear writing is how you prove clear thinking here, because your writing becomes the spec an agent builds from.

How you work. You treat coding agents as leverage you verify, not staff you delegate to and trust blindly. You can still do the underlying work by hand, and that is exactly what lets you catch an agent's mistake before it ships. You move between a strategic question and an implementable spec inside a single day without getting stuck at either altitude; there is no team here to carry you between the two.

What you've probably built. You have owned a product line or a major surface end to end and can point to a decision you made and the outcome it produced, not a roadmap deck you presented. Comparable experience counts: a general manager who ran product and go-to-market together, or a founder who built and sold their own product. We weigh the artifact and the reasoning behind it far more than the title on your last job.

Who this isn't for. This is wrong for you if the appeal of "VP" is the org you expect to inherit; there isn't one, and building one is not the job. It's wrong if you need a prior brand name on your resume to feel safe taking the next step, or if your instinct in a new domain is to defer to whoever seems most confident in the room rather than build your own model of what's true. It's wrong if you want one lane and a clear boundary around it; this seat's boundary is the whole product loop. It's wrong if 'VP' on your resume for eighteen months is the actual goal; this seat is a multi-year commitment in both directions. And it's wrong if you'd rather manage the person who writes the spec than be the person who writes it. You'll be happiest here if the thing you actually want is the scope and the economics of ownership, with or without the title that usually comes attached to a team.

How We Evaluate

We don't run traditional executive interviews.

1. Async video screen. About 15 minutes, on your own time. It replaces the recruiter screen. We want to see how you think, not how you present.
2. Calls with company stakeholders. Short conversations with the people you'd actually work beside.
3. Conversation with the founder. How you model the system above, where you push back, and whether you can hold an argument live instead of retreating to a deck.
4. Paid work trial. Real work in the real environment, taking a live product problem from intent to locked spec to verified build. We watch whether you get grounded fast, whether you write the spec before the build, how you verify what an agent hands you back, and whether your own self-assessment is honest.

If the work above reads like yours but your resume is unconventional, apply anyway. We hire on the work and the reasoning, not the pedigree.

Compensation & Ownership

Total first-year comp: $325,000 - $475,000 (base + performance-based ownership and profit-share programs).

Base: $250,000 - $300,000.

Beyond base: eligibility for the company's ownership and profit-share programs; grants are performance-based, terms discussed at the offer stage. 100% family premium coverage. An effectively unlimited token budget, steered by return, never capped.

Based in Santa Monica, Los Angeles, in person, five days a week. Relocation support available for the right builder.