Staff Research Engineer, Model Efficiency

About Cohere

Cohere is a leading enterprise AI firm focused on security-first foundation AI models and comprehensive products addressing practical business challenges. We develop frontier large language models (LLMs) to empower enterprise AI systems and contribute to accelerating AI adoption globally. Our diverse team comprises researchers, engineers, and designers committed to enhancing our models' capabilities and impact.

Role Overview

The Model Efficiency team focuses on advancing and maximizing inference efficiency for LLMs across our foundation models. This includes innovating in model architecture, mixture-of-experts (MoE) routing, decoding and inference algorithms, and software-hardware co-design to leverage GPU acceleration without compromising model performance.

While based in Toronto, Montreal, San Francisco, New York, Paris, Seoul, and London, the team operates remotely with a preference for Eastern and Pacific time zones for collaboration and flexibility.

Key Responsibilities

• Design, prototype, and implement methods to significantly enhance the runtime speed and efficiency of models in production environments.

Qualifications

• PhD in Machine Learning or a related discipline.
• Deep knowledge of LLM architectures and strategies to optimize inference under resource constraints.
• Demonstrated expertise with one or more state-of-the-art techniques improving model efficiency.
• Strong programming and software engineering capabilities.
• Ability to thrive in a fast-paced, high-ambiguity startup setting.
• Published research in leading conferences such as ICLR, ACL, or NeurIPS.
• Enthusiasm for mentoring and supporting colleagues.

Location & Work Environment

This position supports remote work or office presence in various cities globally without minimum office attendance requirements. Preference is given for candidates aligning with EST and PST time zones for team synchronization.

Perks & Benefits

• Weekly lunch stipend equivalent to $75 in local currency.
• Comprehensive health and dental coverage including dedicated mental health funds.
• Retirement benefits such as RRSP matching, 401K, or pension schemes.
• Full parental leave pay top-up usable up to six months for either parent.
• Annual enrichment allowances for arts, culture, fitness, wellness, quality time, and workspace improvements.
• Education and learning stipend for conferences, courses, and coaching.
• Generous 6 weeks (30 days) of paid vacation.
• Travel budget for office visits and yearly company offsite events.
• Support for remote workers including coworking benefits and a $500 home office setup allowance.

Diversity and Inclusion

Cohere champions an inclusive workplace welcoming applicants from all backgrounds and offers accommodation upon request throughout the recruitment process.

Additional Information

To ensure integrity, Cohere uses AI-enabled tools for candidate screening but does not exclusively rely on them. Beware of fraudulent communications posing as Cohere recruiters; genuine offers come from official company domains.

Compensation

The salary range for this role is CAD 250,000 to CAD 535,000 annually.