The moment

Physical AI is leaving the lab. Frontier models are meeting real machines in real buildings, doing real work — and almost nobody has figured out what happens next. Fleets that need to be watched. Autonomy that needs to be bounded. Knowledge that needs to compound instead of evaporating when an engineer leaves.

Formant is the operating layer underneath all of it. We are the neutral convener: the place frontier AI labs and hardware powerhouses meet, and the system their physical agents run on in production.

We are roughly 15 people. The category is being written in real time by a small number of teams. We intend to be one of them.

The role

This is not a seat on a machine someone else built. There is no playbook to inherit. You will write one.

In your first six months, you will:

• Open doors that don't exist yet. Find the labs, integrators, OEMs, and operators building physical AI, and get into the room. Cold, warm, conference floor, group chat — whatever it takes. Most of these conversations will go nowhere. You will keep making them.
• Turn conversations into partnerships. Take an early signal and carry it all the way to a signed, shipped, referenceable deployment. Not a logo slide — a working thing in the world.
• Find the shape of the market. Notice which pitch lands and which dies. Write down why. Change what we say next week. Repeat until it's a repeatable motion, then hand it to the next person we hire.

Who this is for

The line between technical people and go-to-market people is dissolving. We don't care which side you started on. We need a full stack GTM hacker.

• You may have been a founder — you raised, you sold, you shipped, you did the taxes.
• You may be a business person who is unafraid to read the docs, run the SDK, and hold your own with a robotics engineer.
• You may be an engineer who is unafraid to walk onto a factory floor, into a conference hall, or into a room of strangers and start connecting.

What matters is that you can hold both and expand into whatever the company needs next quarter.

You should be:

• A generalist by conviction, not by accident. You've deliberately built range because narrow roles bore you.
• Genuinely fired up about physical AI. Not "AI is hot." You have opinions about what happens when models get bodies.
• Comfortable with ambiguity that never fully resolves. Some weeks the priority changes on Wednesday.
• Willing to knock on every door. Rejection is the job, not a signal that something has gone wrong.
• Honest under pressure. We would rather hear the bad number early than the good story late.

Preferred experience

• Experience in sales, partnerships, or account management of a highly technical product — at a B2B SaaS, frontier AI, physical AI startup, or data collection company.
• Experience working with both startups and enterprise companies, and comfort navigating both fast and longer sales cycles.
• Experience negotiating and closing deals from all sizes

Who this isn't for

• Someone who needs a defined territory, a quota model, and a marketing team feeding them leads. That infrastructure does not exist here. You are the infrastructure.

Practicals

• Location: Bay Area, hybrid — 2 days in office / remote with travel expectations
• Compensation: $145k–$200k base + variable, plus meaningful equity
• Travel: Regular. Conferences, customer sites, partner labs.
• Reports to: Anne Puyt, VP of GTM

How to apply

Skip the cover letter. Send us:

1. Your CV or LinkedIn.
2. A short reasoned note - a few paragraphs, not a deck - on one company building in physical AI that you think should partner with Formant, and how you'd get the first conversation with them.

We read every one. If the note is good, we'll talk within a week.

Formant is an equal opportunity employer. We hire for range, judgment, and appetite — and we know those show up in people whose résumés don't look alike.