Applied Research - Scientific AI

About the job

Responsibilities

• Build long-horizon environments for agents to search literature, query scientific databases, run code and domain tools, and choose experiments within a budget.
• Translate research partners' workflows, datasets, and outcome criteria into trainable tasks, evaluations, and reward signals.
• Design verifiers for correctness, novelty, and reproducibility, checking for unsupported conclusions and label leakage.
• Connect programmatic checks, simulations, domain models, and expert feedback to training, accounting for noisy results and defining needs for human review.
• Develop RL and multi-agent methods for hypothesis proposal, testing, and critique, comparing them with single-agent baselines.
• Ensure code, data, logs, and experiment configurations are reproducible, re-running promising results and using ablations to isolate model improvements.
• Release reusable environments, evaluations, and research code.

Requirements

• Depth in RL, agents, or post-training systems, with the ability to acquire relevant scientific context.
• Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
• Strong experimental judgment, including meaningful baselines, careful trace inspection, and clear failure attribution.
• Strong Python skills and the ability to build runnable, inspectable, and extensible artifacts for other researchers and engineers.
• Comfort owning ambiguous problems and keeping partners and teammates updated.

Benefits

• Cash Compensation Range of $150-300k
• Equity incentives
• Flexible Work (San Francisco, New York, London, Berlin, or remote)
• Visa Sponsorship & relocation support
• Professional Development budget
• Team Off-sites & conference attendance