Founding AI Engineer & Entrepreneur in Residence
$90k – $150k 6.0% – 15.0%
New York City
3 years of exp
Full Time
About the job
The short version
- You join as the founding engineer at a vertical AI studio in New York, working directly with a founder who has already built a category leader in a legacy industry.
- You build the first product end to end, from customer calls to production.
- When it's ready to stand on its own, you spin it out as co-founder, with real equity, capital, and a first wave of customers behind you.
About Vertex
Most of the real economy still runs on clipboards, fax machines, and a guy named Dave who knows where everything is. Freight docks. Rental yards. Hangars. Substations. Trillions in revenue, and the software hasn't meaningfully changed since dial-up.
Vertex builds AI-native companies for those industries. We pick a vertical, go deep, build the product the industry should have had, and stay until it wins. We don't advise, we operate. We don't flip, we compound.
The founder has done this once before: a single idea in a legacy industry, scaled to hundreds of employees, tens of thousands of businesses on the platform, operations in multiple countries, and tens of millions raised. Vertex is that playbook, run again. You're hire number one.
The role
This is two jobs in sequence.
Months 1–12: Founding Engineer. You and the founder pick apart one vertical (already chosen, we'll tell you on the first call). You sit in on customer calls, then build. Full-stack ownership of v1: data model, LLM/agent layer, evals, UI, deploy. You'll ship to real paying customers inside the first quarter.
Month 12+: Co-founder. The product becomes its own company. You lead it as co-founder and technical lead, with Vertex as your capital, first customers, and operating partner. Your equity is in that company, not in a holding entity.
You report to the founder. There's no one between you.
What you'll do
- Own the product from zero: architecture, stack, and the first ten thousand lines
- Build production LLM features that survive contact with non-technical users in the field: retrieval, agents, structured extraction, evals, guardrails
- Turn messy, decades-old workflows and data (PDFs, spreadsheets, legacy ERPs, phone calls) into something a model can act on
- Talk to customers every week and let what they say change what you build
- Hire the first two or three engineers once the spinout is funded
- Set the engineering culture for a company that doesn't exist yet
What you've done
- 3+ years as a software engineer, with real ownership of at least one product from 0→1
- Shipped production AI that real customers used: LLM apps, agents, RAG, fine-tuning, or similar. You've debugged a model in prod, not just a demo
- Full-stack fluency: Python or TypeScript, a modern web framework, Postgres, cloud infra, CI. You can run the whole thing yourself for a while
- Comfortable with evals as a first-class engineering problem, not an afterthought
- Can explain a technical decision to a mechanic, a dispatcher, or a CFO and have them nod
- Based in New York or willing to relocate now. We work in person
Nice to have
- Time inside a industry: logistics, construction, aviation, energy, industrial services. Even a summer job counts
- Founder, founding engineer, or first-ten employee somewhere before
- Shipped something people paid for, on your own
- Opinions about tooling, formed by being burned
What you get
- Market salary from day one. This isn't a ramen-equity pitch
- Co-founder equity in the spinout, with standard four-year vesting and a one-year cliff
- A founder who has scaled this exact kind of company sitting next to you, not advising from a board seat
- Capital and distribution already lined up for the first product
- A seat in New York with a small team that would rather ship than talk
Not for you if
- You want a spec handed to you
- You've never talked to a customer and would rather not start
- You need a large team around you to feel productive
- Legacy industry sounds like a downgrade instead of an opening