Bridge AI research and production through model post-training at Cohere, using Python, JAX, and supercompute infrastructure. Work remotely or in global offices.
As a Member of Technical Staff, you will contribute to research and production code according to organizational needs. You will improve model performance, train large models at scale, and collaborate with specialist teams. The role includes experimentation across SFT and RL.
Key Responsibilities:
- Advance model post-training techniques
- Write high-performing software for training models
- Optimize performance across SFT and RL
- Coordinate models with Agentic and Code teams
- Implement and test research ideas on infrastructure
Requirements:
- Extremely solid software engineering skills
- Proficiency in Python, JAX, PyTorch, and XLA/MLIR
- Distributed infrastructure experience with Kubernetes, Slurm, or Ray
- Experience training large models at scale
- Hands-on post-training performance optimization experience
Apply research engineering and model optimization skills to strengthen Cohere’s production systems and frontier model training work.
📌 AI Research Engineer Model Post Training (Canada)
🏢 Cohere
📍 Canada