Member of the Technical Staff, Molecular Generation

📋 Description

• Design and build generative architectures for molecular data across modalities.
• Develop training methods that learn from diverse signals to create novel structures.
• Enable controllable, targeted generation to meet properties and constraints.
• Integrate biological reasoning from our foundation model to guide generation.
• Own end-to-end training: design experiments, multi-GPU training, hyperparameter tuning.
• Design evaluation frameworks to measure biological meaning, validity, and novelty.

🎯 Requirements

• PhD in CS/ML/Physics/Math or related; 2+ years post-doc or 5+ years hands-on.
• Strong publication record in generative methods (NeurIPS, ICML, ICLR).
• Extensive experience designing, building, and training deep generative models.
• Proficient in Python and PyTorch; experience with distributed multi-GPU training.
• Own full research-to-training pipeline; train and ship models.
• Production-quality, well-tested code; comfortable with version control and code reviews.

🎁 Benefits

• We encourage new ideas, creativity and contrarian thinking.
• Healthy, feedback-focused environment with constructive feedback and growth.
• You own your day-to-day management; we hit milestones together.
• Competitive salary and equity in a growing startup.
• Excellent medical, dental, and vision coverage.
• Supportive leadership and opportunities for growth.