Founding RL Engineer

Our client is a venture-backed, seed-stage AI data company based in San Francisco. They supply high-quality multilingual voice data and enterprise data to leading AI labs, and use real-world company data to build rich reinforcement learning (RL) environments rather than synthetic ones. The business already generates meaningful revenue and works with top voice and frontier AI labs. The team is small, founder-led and hiring fast.

We are looking for a Founding RL Engineer to build high-quality RL environments and AI agents. The role spans two streams: building RL environments from the company's proprietary voice and enterprise datasets, and building and deploying AI agents for enterprise customers.

This is the first hire on the engineering team. You will report to and work directly with both co-founders, and lead the engineering function as the team grows.

• Build high-quality RL environments from proprietary voice and enterprise datasets
• Build and deploy AI agents for enterprise customers
• Partner with the co-founders on technical direction and roadmap
• Lead the engineering team as it grows
• Ship fast while holding a high-quality standard at startup pace

• 3-7 years of engineering experience with strong RL fundamentals
• Prior experience building RL environments, either at an RL company or at a human data company with hands-on RL exposure (e.g. Surge, Mercor, Scale-type data companies, Proximal, or comparable)
• Hands-on experience building AI agents in production
• Strong AI-native tooling fluency, using tools such as Claude Code or Cursor daily as production aids
• Based in or willing to relocate to San Francisco, working fully in person 5.5 days a week

• Prior experience at a top human data company
• Full-stack engineering background (TypeScript, AWS)
• iOS, Swift, or React Native experience for mobile app development
• Prior startup experience with a demonstrated shipping cadence

• Location: San Francisco, CA (fully in person, 5.5-day week, Sundays off)
• Employment: Full-time
• Experience: 3-7 years
• Visa sponsorship: Not available

Intro chat with a co-founder, chat with the second co-founder, take-home assignment, technical chat, and a short work trial (a couple of days). The full process can be completed in about one week.