About the Role
Polytope Bio, an Astera project, is developing a post-training engine to connect AI models with high-throughput biology, enabling generative biology applications by aligning AI models with experimental measurements. We seek a Senior/Staff AI Research Scientist to influence modeling, experimental design, and training strategy, working with significant compute and resources to develop and publish new reinforcement learning and generative AI methods.
Responsibilities
- Drive core research by developing and executing the scientific vision and cutting-edge modeling approaches.
- Design, implement, and improve model post-training methods for biological language models using RL feedback loops.
- Architect model training infrastructure and build, run, and debug models, training loops, and evaluation metrics.
- Partner with the experimental team to integrate laboratory measurements and model learning into a cohesive system.
Requirements
- PhD in machine learning, computational biology, or related field, with 1-2 years post-PhD research/industry experience (or equivalent research accomplishments).
- Experience training models from scratch, including debugging training runs.
- Hands-on experience with generative diffusion models and/or transformer architectures.
- Familiarity with modern reinforcement learning and preference-optimization methods for deep learning.
- Demonstrated track record of strong research through publications, open-source work, or shipped models.
- Self-directed with a collaborative mindset, comfortable in an early-stage environment.
- Comfort operating with ambiguity and a desire to build new initiatives.
- Familiarity with biological research (protein modeling, sequence models, structural biology, or adjacent areas) is a plus.
- Experience building and scaling training infrastructure on large GPU clusters is a plus.
Skills
- Machine Learning
- Computational Biology
- Generative Diffusion Models
- Transformer Architectures
- Reinforcement Learning
- Preference Optimization
- Deep Learning
- Model Training
- Research
- Scientific Vision
- Experimental Design
- Training Strategy
- Infrastructure Architecture
- Debugging
- Collaboration
- Open-Source Development
Location
- NYC
- SF Bay Area
- Remote
Work Type
- Full-time
- Residency
- Potential for Co-founding Role
Experience Level
- Senior
- Staff
- 1-2 years post-PhD research or industry experience
Education Level
- PhD in machine learning, computational biology, or a related field
Salary/Compensations
- $200,000 to $300,000 during the residency
Benefits
- Full benefits package including health insurance
- Company sponsored retirement plan
- Vision
- Dental
- Authorship of high-impact open-source datasets, methods, and models
- Significant secured runway
- Dedicated GPU resources
About the Company
- Astera is a private foundation dedicated to advancing science and technology for an abundant future.
- We support projects that operate like high-velocity startups, focusing on ambitious goals and attracting top talent.
- Polytope Bio is a residency project at Astera focused on closing the loop between AI models and high-throughput biology.
