Machine Learning Scientist — Large Multimodal Models (Post-Training)

Highlights

• Develop post-training methods for Enchant, a multimodal biomedical foundation model.
• Put powerful foundation models in hands of scientists making therapeutic decisions.
• Design reward functions and RL strategies to push drug-discovery capabilities.

What they offer

• Company-paid healthcare
• 401(k) matching
• Uncapped vacation
• Flexible spending accounts
• Voluntary life insurance

Ideal candidate

A curious researcher with a PhD mindset who loves rigorous experimentation and wants foundation models to drive real therapeutic breakthroughs, thriving in close collaboration with scientists and engineers.

Requirements

• PhD in machine learning, computer science, computational chemistry, physics, or related computational STEM field, or equivalent industry experience
• Strong Python and PyTorch skills, including implementing, training, debugging and evaluating deep learning models end-to-end
• Demonstrated experience training large-scale transformer models
• Demonstrated experience in reinforcement learning such as RLHF, RLAIF, PPO, GRPO, RL with verifiable rewards (strongly preferred)
• Experience with supervised fine-tuning, full-parameter fine-tuning, parameter-efficient fine-tuning (LoRA), or related methods
• Experience with systematic hyperparameter optimization or large-scale experimentation using Optuna, Ray Tune, or similar
• Strong engineering practices: reproducible experimentation, clean code, testing, and performance-aware debugging
• Comfort with modern ML infrastructure such as Docker, CUDA, Kubernetes and Weights & Biases