Applied AI Engineer - Long-Horizon Agents

What if an agent didn't stop after one task, but kept working on your goal for weeks, months, or indefinitely?

That's the bet behind a brand-new AI lab in San Francisco. The founding team comes from leading frontier AI labs, autonomous-driving and robotics AI, big-tech research and a top quant firm. Their thesis: once agents run long enough, raw capability stops being the hard part. Knowing what the human actually wants is. So they're building agents that take sparse direction and messy evidence gathered over time, form a lasting picture of the goal, and then plan, act, check outcomes, recover from mistakes and keep improving. The first application: products that improve themselves, where agents learn what customers are trying to get done, build and ship the solution, measure it and iterate.

They're hiring 4 founding-level engineers.

What you'll own

• Whole product areas, zero to one, end to end
• The surface where users hand complex outcomes to agents and stay in control
• The full stack: UI, agent workflows, model orchestration and the infra underneath
• Turning what frontier models can do into products that hold up for real users, not demos
• Time with users: learning their workflows and shipping iterations fast
• Shaping product, architecture and culture as one of the first engineers

What you bring

• World-class engineering skills, combined with a sharp eye for product and real taste
• A track record of shipping products people love, ideally from zero
• Sharp instincts for turning new model capabilities into something simple and useful
• Range across product, engineering, models and infra: a builder, not a narrow specialist
• Proof of serious ownership: you started a company, ran a major product, or shipped work that mattered at a frontier-tech company
• Roughly 3–6 years in; exceptional outliers and senior leads are welcome too
• High agency, high speed, founder mindset. You don't need hand-holding.

Bonus points

• You've built at a top AI coding or agent company
• Founder or early-team experience (an exit is a strong signal)
• A spike: top of your field in olympiads (IOI/IMO), competitive programming, chess, sports, or quant trading

What's in it for you

• $250k–$500k base + 1–5% equity
• Founding-level ownership, alongside an exceptionally strong founding team
• Open, hard problems at the frontier of long-running autonomous agents
• Visa sponsorship available; if you're outside the US, expect to go via an O-1

Good to know

• Full-time, in person in San Francisco, 9-9-6. Intense and high-ownership by design.
• Process: informal talk with a founder - technical deep-dive on your applied-AI work - paid 2–3 day work trial in person (expenses covered) - offer

Does this sound like you?