Founding AI Engineer
About the role
A pre-seed AI startup is hiring a Founding AI Engineer to build and lead the core of its product: an AI voice agent that sells. The agent joins live calls with prospects, asks discovery questions, shares its screen and demos the product by controlling a real browser, then follows up until the deal closes. The company is live with paying customers and has more demand than it can handle. You would join as one of the first engineers on a small team, report directly to the CTO and co-founder, and own the agent technical direction, including architecture, models and how the team proves it is getting better. Offers come in two versions: more cash with less equity, or less cash with more equity.
What you will do
- Agent capabilities (about 40%): build the reasoning, tool use and memory of the agent, and how it recovers when a live call goes off script, all under tight real-time latency limits
- Evals and data (about 25%): turn thousands of real sales calls into measurable weekly improvements through eval pipelines, data work and fine-tuning where it pays off
- Browser agent reliability (about 20%): make the agent dependable at running a live product demo in a real browser in front of a buyer while it talks
- Shipping across the stack (about 15%): ship end to end in TypeScript, Node and Next.js, and work directly with customers and design partners to shape the product
- Set direction: decide which models and methods the company bets on, what to build versus buy, and what the agent can do next quarter that it cannot do today
What I am looking for
- 4 to 7 years of experience as an AI engineer
- Has owned and shipped AI agents to production from start to finish
- AI-native and opinionated, comfortable making decisions independently without a PM layer
- Excited by equity and the ride rather than stability and perks
- Able to work from the New York office 4 days a week
About the company
An AI startup building voice agents that act as full-cycle sellers for B2B software companies: qualifying prospects, answering complex questions, demoing the product live in a browser, and following up until deals close, around the clock. It was founded by serial entrepreneurs who previously sold their first company. It recently closed a pre-seed round backed by well-known founders and institutional investors, and it has just started generating revenue, with demand well beyond what it can currently serve. It is a founder-led team of about 8 where everyone has high ownership and works directly with customers, and the team uses AI tools heavily to operate like a much larger organization.