Senior Multimodal AI Systems Architect

Cohere; United States; On-site; USD 180,000 - 240,000; Full time

Benefits offered by this job

Six weeks vacation

Equity/Stock options

Insurance coverage for health, dental,

Mental health coverage

Parental leave (6 months)

Egg freezing/IVF support (Canada/UK)

Wellness allowance

Remote-friendly culture

Education benefit $2000/year

Meals stipend and catered lunches

Arts and culture allowance

Quality time allowance

Job summary

Cohere in the United States is seeking talented engineers to design and develop cutting‑edge multimodal AI systems, integrating text, speech, and vision. You will conduct research on advanced compute infrastructure and collaborate with world‑class teams to push the boundaries of scalable AI.

We value exceptional software engineering, distributed training expertise, and a track record in multimodal research. This role offers remote work with global teams and generous benefits.

Qualifications

• Exceptional software engineering skills with a track record of building robust, scalable systems.
• Knowledge of distributed training strategies for large-scale multimodal models.
• Publications in top-tier venues demonstrating multimodal AI research (bonus).
• Experience writing efficient GPU kernels using CUDA for multimodal tasks (bonus).
• Familiarity with autoregressive models for multimodal tasks such as image/video captioning and speech-to-text.
• Strong command of Python and popular DL frameworks (JAX, PyTorch, TensorFlow) and multimodal capabilities.
• Passion for ML and its potential to impact various industries through multimodal applications.
• Experience tuning large multimodal models and building evaluations.

Skills

Software engineering

Multimodal models

Python

CUDA

PyTorch

JAX

TensorFlow

Distributed training

Tools

CUDA

PyTorch

TensorFlow

JAX

Job description

Cohere in the United States is seeking talented engineers to design and develop cutting‑edge multimodal AI systems, integrating text, speech, and vision. You will conduct research on advanced compute infrastructure and collaborate with world‑class teams to push the boundaries of scalable AI.

We value exceptional software engineering, distributed training expertise, and a track record in multimodal research. This role offers remote work with global teams and generous benefits.