Staff Machine Learning Scientist
Staff Machine Learning Scientist
A cell-therapy biotech is building a next-generation platform that combines AI-driven foundation models with lab-in-the-loop experimentation to accelerate regenerative medicine. With over $50M in recent funding and a mission to transform how cell therapies are developed, the company leverages vast in-house datasets and cutting-edge machine learning to enable precise control over cell behavior. This is with the aim of unlocking new treatments for complex diseases such as diabetes, neurodegenerative disorders, and muscular dystrophies.
The Staff Machine Learning Scientist will lead efforts to design and optimize foundational models for biological data, collaborating closely with computational biologists and experimental teams. This role combines technical leadership, hands-on research, and infrastructure development to push the boundaries of AI in life sciences.
Key Responsibilities
- Develop large-scale foundation models for high-dimensional biological data, including single-cell RNA sequencing.
- Work at the intersection of deep learning, computational biology, and experimental feedback loops.
- Engineer and scale distributed training and inference systems capable of handling trillion-token biological datasets efficiently.
- Develop and experiment with advanced deep learning architectures, including transformer-based models specifically adapted for complex biological datasets.
- Join a team of top-tier AI researchers and life science experts in a fast-paced, mission-driven environment.
Qualifications
3+ years of experience in machine learning, with expertise in transformer architectures and large-scale model training.
Master's or Ph.D. in Computer Science, AI, Computational Biology, or related field.
Proven track record of end-to-end model development, from research to production. This is in addition to proven independent research
Strong proficiency in Python and deep learning frameworks (PyTorch or TensorFlow).
Bonus: Experience with reinforcement learning (RLHF, PPO), diffusion models, or multimodal architectures.