Member of Technical Staff, Agent Code

About Cohere

Cohere is a premier enterprise AI company prioritizing security. Our mission centers on developing advanced foundational AI models and comprehensive products designed to address tangible business challenges. Operating globally from Toronto, with significant offices in London, New York City, San Francisco, Montreal, Paris, Berlin, and Seoul, we emphasize innovation and value creation in AI deployment.

Role Summary

Join our pioneering Agent Code team as we drive transformative advancements in code-generating large language models (LLMs) and autonomous agents that redefine software creation and task automation. This senior role combines research and engineering to develop scalable code LLMs and agent systems capable of interacting with digital environments to tackle complex tasks with limited human input.

Key Responsibilities

• Continuously monitor and incorporate cutting-edge research on code LLMs and agent technologies into practical implementations.
• Architect and deploy scalable training methods for code models and production-ready agent platforms, collaborating closely with pretraining teams, and engaging in supervised fine-tuning (SFT) and reinforcement learning (RL) algorithm development.
• Optimize existing benchmarks and design innovative evaluation metrics that align with enterprise user needs.
• Lead experimental initiatives utilizing state-of-the-art computational infrastructure to push the performance frontier of LLMs.

Required Qualifications

• PhD in Computer Science, Machine Learning, or an allied discipline with published work in premier conferences (e.g., NeurIPS, ICML, ICLR, ACL, EMNLP).
• Proven deep expertise in code LLMs and agent frameworks, including active contributions to their research or development.
• Practical experience operating frontier LLMs with applications in code generation or automation contexts.
• Strong programming and software engineering proficiency, particularly in Python and machine learning frameworks such as PyTorch or TensorFlow.
• Familiarity with distributed systems, cloud computing platforms, and scalable system architectures.
• A proactive and self-directed approach, driven by tackling challenging, open-ended problems.

Work Environment & Location

This position supports remote flexibility but favors candidates within Eastern to Central European time zones to foster team collaboration. Offices are located in London, Toronto, New York, and San Francisco. There is no mandatory on-site presence requirement.

Employee Benefits

• Weekly meal allowance equivalent to $75 USD or £75 GBP to support lunch expenses.
• Comprehensive health and dental insurance, including dedicated mental health support budgets.
• Retirement benefits with RRSP matching, 401(k), or equivalent pension schemes.
• Generous parental leave policy offering up to six months with full pay top-up for either parent.
• Annual enrichment allowances covering arts, culture, fitness, wellness, workspace improvements, professional development including conferences and coaching.
• Six weeks (30 working days) of paid vacation annually.
• Travel budgets for remote employees to visit other offices and participation in yearly company-wide offsite events.

Additional Information

• Cohere fosters a diverse and inclusive workplace, welcoming individuals from various backgrounds and providing accommodations throughout the recruitment process if needed.
• Applicants should beware of scams; all legitimate job listings are posted only on official company platforms with email communications from @cohere.com or @cw.cohere domains.
• Recruitment processes may incorporate AI-assisted screening tools to help identify qualified applicants without limiting human assessment.
• The compensation range for this position is between CA$250,000 and CA$535,000 annually.