Research Engineer LLM Performance
AI-first drug design and development company building a unified drug-design engine to advance new medicines.
Funding history
Investors
About Isomorphic Labs
Isomorphic Labs uses predictive and generative AI models, computational biology, and drug-design expertise to develop therapeutics across partnered and internal programs.
Skills
Candidate Availability
Required and preferred rules are kept separate and reflect the wording in the original posting.
About the Role
You will implement and optimise LLM post-training methods on frontier models. You will turn research methods into production-ready systems, prioritise performance improvements, evaluate fine-tuning and reinforcement-learning frameworks, diagnose distributed-system bottlenecks, and deploy low-precision methods that balance model performance and accuracy.
Requirements
- Significant experience with large-scale distributed LLM training
- Experience with JAX or PyTorch
- Knowledge of parallelism strategies and collective communication libraries such as NCCL
- Understanding of GPU architectures
- Excellent collaboration skills
Responsibilities
- Implement and optimise LLM post-training methods at scale
- Translate research methods into production-ready systems
- Identify and execute performance optimisation opportunities
- Evaluate and deploy supervised fine-tuning, reinforcement learning, and LLM evaluation frameworks
- Diagnose and fix bottlenecks in distributed training and inference
- Deploy low-precision methods that balance performance and accuracy