Member of Technical Staff Safety for Agents

@ Cohere

About Cohere

Our mission is to scale intelligence to serve humanity. We build frontier AI models for developers and enterprises in areas like content generation, semantic search, RAG, and agents. We work hard, move fast, and value diverse perspectives to create great products.

Role Overview

As a Member of Technical Staff Safety for Agents, you will focus on data generation, post-training algorithms, and evaluation methods to enhance safety in Large Language Models (LLMs). You will collaborate with cross-functional teams including machine learning, data annotation, product, and policy teams to build better, fairer AI systems.

Responsibilities

• Design and conduct data collection tasks with human annotators.
• Implement safety algorithms and evaluate model performance.
• Collaborate with cross-functional teams to improve model robustness.
• Analyze datasets for quality, biases, and training suitability.
• Contribute to scientific research and publish findings.

Required Skills and Experience

• Strong statistical and experimental design expertise.
• Excellent software engineering skills.
• Experience training large language models on distributed systems.
• Proficiency in Python and ML frameworks like PyTorch, TensorFlow, or JAX.
• Outstanding communication and collaboration skills.
• Publications in top-tier conferences are a plus.

Perks & Benefits

• Inclusive and open work culture.
• Access to cutting edge AI research and technology.
• Weekly lunch stipend, health and dental benefits.
• Remote-flexible work with several global office options.
• 6 weeks vacation and enrichment benefits.

Key skills/competency

LLMs, data generation, safety algorithms, Python, ML frameworks, distributed training, experimental design, data analysis, research, collaboration.