About the RoleThis is a true co-founder seat, not a senior hire, at an early-stage creative AI and marketing technology company with a live product already generating rich preference data from real brand campaigns. You will own the entire technical and research vision, from today's creator workflows to a proprietary training and evaluation platform for creative AI agents. The role exists because the data and research opportunity here is unlike anything available at larger organizations.
What You'll Do- Own technical and research strategy, defining the path from current workflows to a proprietary training and evaluation platform.
- Build data infrastructure that connects briefs, submissions, generation traces, decisions, critique, and campaign outcomes.
- Design systems that evaluate, rank, critique, generate, and improve creative work.
- Develop the allocation policy that directs campaign budget and simultaneously shapes dataset collection.
- Make core decisions on frontier model usage, fine-tuning, post-training, reinforcement learning, and original research directions.
- Set standards for data quality, consent, privacy, measurement, and model evaluation.
- Recruit and lead the first engineering and research team.
- Work directly with creators, brands, and investors as a company representative.
What We're Looking For- Hands-on experience building preference or reward models, including RLHF, reward modeling, ranking subjective outputs, or evaluation systems.
- Strong grounding in causal inference on observational data: selection bias, off-policy or counterfactual evaluation, adaptive experimentation, or validating a cheap proxy against an expensive delayed outcome.
- Production ML at scale in ads, marketplace, recommender, or ranking systems tied to real business outcomes.
- Experience with bandits, sequential allocation, or online experimentation in production.
- Track record shipping real products or features used by real people, not research that stayed in papers or prototypes.
- Background at a frontier AI lab or research organization of comparable caliber is a strong positive signal.
- Experience with multimodal generation, agents, planning systems, or latent-variable modeling is valued.
- Proven ability to recruit and lead strong technical talent and own programs end to end.
- Prior founder experience is a plus.
- Ready to commit fully (6 to 7 days per week) as an equal co-founder, comfortable with early-stage uncertainty and full ownership of technical strategy.
Compensation & Benefits- Salary: $250,000 to $500,000 annually, discussed based on circumstances and fit.
- Equity: Approximately 28% as an equal co-founder.
- Visa sponsorship: Not available.
LocationOn-site, full-time in San Francisco, CA.
