AI-Native Product Manager

The founder's product execution partner. You take a surface from strategy to shipped - spec, design direction, agent-built implementation, growth loop - without it changing hands.

Product.ai is the verified truth layer for shopping: when a person or an AI agent needs to know what is actually true about a purchase, we answer with proof. SimplyCodes is the first proof at scale, the code verification service whose robots run real checkouts so shoppers only see codes that actually work. It earns about $22 million a year at roughly 60% margins. Profitable. Bootstrapped. Founder-owned since 2009. No outside investors. No board. Fewer than twenty operators, outbuilding companies 10x our size.

Why This Role Exists

The founder runs product strategy personally, with AI agents at every stage of the build. The model works: a small team ships consumer truth surfaces, a conversational shopping expert, and commerce landing systems at a pace small teams are not supposed to reach. What the model does not scale is his hands. Today every surface crosses the founder's desk between strategy and shipped. This seat is the deliberate fix: a product execution partner who takes a surface he has framed and carries it end to end - through spec, design direction, agent-run implementation, and the growth loop that proves it - and hands back something production-ready.

You work directly with the founder, daily. It is the closest product seat to him in the company, and deliberately not the most senior one. We have a Director-altitude product posting open; this is not it. This seat is for the operator one rung earlier who is disproportionately dangerous with AI in their hands.

The System You'll Need to Model

• Probabilistic product management. The product is a truth engine: it answers with confidence, cites its evidence, and refuses when it cannot verify. Managing it is evaluation architecture, not roadmap theater. You define "good" as a rubric a machine can grade, and the eval is the spec.
• The verdict surface class. How verified truth renders to a human: confidence a shopper can feel, refusals that build trust instead of losing the sale, provenance that survives a skeptical reader. The same verdict has to hold up on a web page, in a chat answer, and inside someone else's AI agent.
• Incentive-designed contribution systems. Crowdsourced verification with real rewards - the system class behind every marketplace of human effort. Reputation design, contribution ledgers, anti-gaming checks. The physics you'll inherit: for complex tasks, standing in a community usually beats cash.
• Agent-built production. Agents write much of the code and the content here; humans own design, failure modes, and verdicts. Cost per query is a real design constraint, so you steer token spend by measured return the way an older generation of PM steered headcount.
• High-traffic commerce surfaces. Landing systems measured at the click and the dollar, at hundreds-of-thousands-of-pages scale, where search physics and revenue per click are the scoreboard and every change ships behind a pre-registered test.
• Cortex, the shared AI brain. The governed substrate the company runs on and the product family we sell - it answers its own questions from more than 8,600 documents. Your job runs inside it, not alongside it.

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

What You Will Own

• The founder's product lane. A surface arrives framed; you return it production-ready. You write the spec, set design direction with the founding designer, run the agent-built implementation with the engineers, and close the growth loop that proves it moved a number. The bar: a framed strategy becomes a live surface without returning to the founder's desk for rescue.
• One consumer contribution product, end to end. An incentive-driven verification program - UI-heavy on the front, incentive design and quality control underneath. You own the loop: who contributes, why they come back, and what keeps the signal honest against people trying to game it.
• The experiment layer. Every surface change ships behind a pre-registered test with a written ship-or-kill verdict inside days. You run that discipline and publish the verdicts, good or bad.

The craft you must already own: product management fundamentals - problem selection, spec writing, tradeoff calls, shipping. What you'll grow into here: running agent fleets against product outcomes, designing the evaluations that make an agent's work trustworthy, and directing design and engineering at once without a coordination layer. You partner with the seat that adjudicates the truth layer and the founding designer who owns how verdicts feel; your lane is carrying surfaces to done.

Who You Are

You think end to end. Strategy, design, technology, and growth are one system to you, not four departments to route between. You form a working model of a new system fast, notice where your model is wrong, and update it in public.

AI-native is not a tool preference for you; it is how you work. You have run the agents yourself - you can do the job by hand and prove it, and you direct agents the way our founder does: set the outcome, design the check the agent cannot fake, own the verdict on what comes back. You automate before you delegate. You write clearly, because clear writing is evidence of clear thought.

What you've probably built: a real product system in the AI era, with a number attached. A legacy flow you re-architected into an agentic system and the dollars it moved. A consumer surface you took from spec to live. A marketplace or incentive loop you designed and defended against gaming. We care about the artifact and the reasoning, not where you did it or how many years it took.

Who this isn't for. This seat fits someone who wants maximum surface area with a founder and treats a framed problem as fuel. It is the wrong seat if you optimize for proximity to power instead of owning outcomes. Wrong if your next move is chosen by how the title will read at a bigger brand. Wrong if you need a squad of specialists between you and shipped - a researcher for the research, a designer for every screen, an engineer for every experiment. And wrong if you wait for the task list; here you write it. The right person values the opposite: the full altitude range, the direct line, and a verdict in days instead of quarters.

How We Evaluate

We don't run traditional product-manager interviews.

Async video screen. Brief and self-recorded: about 15 minutes, whenever works for you.

Calls with company stakeholders. Short conversations with the people you'd build beside.

Conversation with the founder. How you model a product system - and what you would ship first.

Paid work trial. Ten to fourteen days of paid work in our real environment, on a real surface, at your stated rate. We say so up front on purpose: paying well for real work respects your time and draws high-agency people. We watch four things - how you ground yourself in our systems, whether you write the spec before the build, how you verify what your agents produce, and whether your 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: $300,000 - $425,000 (base + performance-based ownership and profit-share programs). Base: $200,000 - $260,000 - top of market for senior AI-native product management.

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; and an effectively unlimited token budget, steered by ROI, never capped.

This is a partnership structure, built to mint partners. When the company wins, you win - in real, liquid dollars, every year.

Based in Santa Monica, Los Angeles - in person, five days a week. The rooms are real rooms.