About the role
Cohere is hiring a Member of Technical Staff for Data Analysis and Evaluation, fully remote with no location restrictions.
This role owns the quality, reliability, and performance evaluation of Cohere's frontier LLMs. You design and oversee data collection tasks (including human annotators), develop statistical methods to evaluate dataset quality, analyze model generalisability and robustness, and conduct experiments to identify improvement areas. You also train and fine-tune LLMs on distributed infrastructure.
Required: extremely strong software engineering, strong expertise in designing data collection tasks with human annotators, strong statistical skills, hands-on distributed-LLM-training experience, Python plus ML frameworks (PyTorch, TensorFlow, JAX), publication at top venues (NeurIPS, ICML, ICLR, etc.).
What stands out: the posting explicitly states "no restrictions on where you can be located" — unusual frankness for a US/Canada-HQ AI lab. Cohere has offices in Toronto, NY, SF, London, Paris with co-working stipend for everyone else.
This recap is dataskew's editorial summary, not the company's copy.