Head of Product - Zingage

New York • In person • Reports to the founders

The problem

The hardest problem in applied AI is not making agents smart - models can already schedule a shift. It is making them trusted, task by task, by the people whose jobs they change. After hours, autonomy is granted by default; the alternative is a groggy on-call coordinator. During the day, it is negotiated in real time with a scheduler whose professional identity is at stake. Fully autonomous at 2AM when no one is watching; a trusted teammate at 2PM when everyone is. Almost everyone working on earned autonomy is working on it in demos. Here it has to be solved in production, where the patient is someone's mother - and the stakes are precisely why solving it means something.

What winning means

For the world: the people who care for our parents - an industry of two million caregivers and the agencies that coordinate them - finally get real infrastructure, and the endgame is becoming the rails an entire industry's coordination runs on. For you: the founder problem without the cold start. Product-market fit is found, revenue is real, the strategy is written and board-aligned, and a 2-3x ACV expansion sits inside the existing customer base as agencies move from after-hours to 24/7. You inherit a full quarter of adversarially-verified ground truth about how autonomous agents fail in the real world - and a team that answered it with owners, key results, and a 60-day durability standard. The conversion from product-market fit to durable franchise is yours to run.

The three builds

We hire for spikes. This role is a measurement-systems product leader - someone who builds the instruments, not someone who reads dashboards. You will build three systems, each wired to a number:

• The persona foundation. Today every agency runs a lightly bespoke configuration; you will collapse that into learned persona baselines on a typed configuration menu, so onboarding becomes diagnosis instead of elicitation, and saying no to exception 121 comes with a better alternative attached. → Onboarding speed, retention.
• The context ingress. We believe the new system of record is the system of context: the durable position in the AI era belongs to whoever continuously collects and compounds the operational context - every call, message, exception, and handoff - that agents need to act. You will build that position: the rails on which caregivers, clients, and admin staff communicate - the phone system, the internal comms, and above all the agent-to-human handoff, which must stay inside our perimeter because it is where trust is negotiated and context is richest. → NRR, via the after-hours-to-24/7 expansion.
• The measurable trust system. Experience measurement is the instrument, not the end state - the end state is trust as a number: how much autonomy each customer grants our agents, how every person our agents touch actually feels about them, and whether both are rising. You will build it from revealed preference first - response latency, offer acceptance, opt-outs, frustration signals per thousand interactions - across caregivers, clients, and schedulers, with NPS as the legible summary on top. This system is simultaneously our churn early-warning, our team's ground truth, and the evidence that prices the company. → Retention, and the proof behind everything else.

The scoreboard

NRR, retention, onboarding speed. Three numbers, your name on them, reviewed with the founders monthly. Product initiatives that don't trace to one of the three don't ship.

The offer

Top of market, and we mean it in numbers: $250-300K base plus 0.5-1.0% equity - well above benchmark for this title, because we are hiring a franchise-builder, not a feature-shipper.

New York, in person, at the founders' table. If you have founded a company - or operated with founder-level ownership and are looking for the problem worth the next five years - we would like to meet you.