Tahoe Therapeutics is looking to solve bottlenecks in drug discovery by building the world's largest in vivo single-cell perturbation atlas and using it to train multimodal foundation models. The company aims to find better drugs faster by combining cutting-edge machine learning with biologically relevant datasets.
Requirements
- Solid Engineering and Computer Science fundamentals, ideally with a degree in CS, Math, or equivalent experience.
- Exceptional engineering skills to iterate quickly on data processing and training pipelines
- Experience in building, testing, training, and deploying modern neural network architectures such as Transformers.
- Experience with frameworks like PyTorch, Tensorflow, Keras, JAX and the ability to write hardware optimized code for distributed training
- Experience with distributed deep learning using frameworks such as HF Accelerate, Deepspeed, Composer, TorchTitan, Megatron etc.
- Experience with training large neural networks on multiple compute nodes.
Responsibilities
- Stay at the forefront of ML and computational biology research and rapidly adopt state-of-the-art techniques to train AI virtual cell models on our massive single cell datasets.
- Work together with ML Scientists to develop the next generation of compute optimized models like Tahoe-x1.
Other
- This position requires on-site presence at our South San Francisco office a minimum of three days per week.
- We welcome applicants who require visa sponsorship and provide work authorization support for qualified candidates.
- Unlimited Paid Time Off (PTO).
- Monthly Lunch budget
- One-time Office set up budget
