Post-Training ML Research Engineer — RL & Production

Warp; South San Francisco (CA); Sur place; USD 140 000 - 230 000; Plein temps

Résumé du poste

Warp in California’s South San Francisco area is seeking a post-training research engineer/scientist to turn advanced models into practical production capabilities. You will own the post-training loop, from identifying capability gaps through evaluation, data development, training, deployment, and feedback, with a focus on fast, reproducible experimentation.

You will develop RL training methods, design reward signals, curate training data from model failures and product feedback, and collaborate

Qualifications

• Meaningful results track record with post-training improvements.
• Deep practical experience with reinforcement learning, reward design, rollout systems.
• Data-driven approach with evals and strong baselines.

Responsabilités

• Own the post-training loop from capability gap to evaluation, data development, training, deployment and feedback.
• Build evals grounded in product needs with clear baselines and held-out tests.
• Develop experimentation systems for fast, reproducible hill climbing and result comparison.
• Develop and improve RL training methods, reward signals, rollout generation, and infrastructure.
• Curate and refine training data based on model failures and product feedback.
• Improve model behavior across Warp’s robotics and foundation-model stack with cross-team collaboration.
• Validate gains through ablations and testing beyond training distribution.
• Ship improvements and measure impact in production for next iteration.

Connaissances

Reinforcement learning

Evaluation & ablations

Data-driven experimentation

RL training systems

Experimentation infrastructure

Post-training research

Engineering excellence

ownership of results

Description du poste

Warp in California’s South San Francisco area is seeking a post-training research engineer/scientist to turn advanced models into practical production capabilities. You will own the post-training loop, from identifying capability gaps through evaluation, data development, training, deployment, and feedback, with a focus on fast, reproducible experimentation.

You will develop RL training methods, design reward signals, curate training data from model failures and product feedback, and collaborate