Senior Staff AI Safety Engineer for Agents

Cohere; New York (NY); On-site; USD 180,000 - 240,000; Full time

Benefits offered by this job

Lunch stipend

Health and dental benefits

RRSP/401K matching

Parental leave top-up

Education stipend

Vacation and offsite

Home office stipend

Job summary

Cohere, a leading security-first enterprise AI company, is seeking a Member of Technical Staff in the Safety for Agents team to advance safer LLMs through data generation, post-training algorithms, and evaluation methods.

You'll collaborate with cross-functional ML teams, work with data annotation, and tackle new scientific problems with autonomy and impact across global offices, while remote-friendly policies support flexible work arrangements.

Qualifications

• Strong statistical skills for evaluating experiments in data collection.
• Excellent software engineering abilities.
• Experience designing data collection tasks with annotators.
• Experience analyzing datasets for quality and biases.
• Hands-on experience training LLMs on distributed infrastructure.
• Familiarity with generalization and robustness of ML systems.
• Proficiency in Python and ML frameworks.

Responsibilities

• Contribute to data generation and post-training evaluation for safer LLMs.
• Develop algorithms to improve safety and trustworthiness of models.
• Collaborate with cross-functional teams including product and policy.
• Work on experiments and datasets to test solutions at scale.

Skills

Python

PyTorch

TensorFlow

JAX

LLMs

Distributed training

Data generation

Experiment design

Communication

Cross-functional collaboration

Job description

Cohere, a leading security-first enterprise AI company, is seeking a Member of Technical Staff in the Safety for Agents team to advance safer LLMs through data generation, post-training algorithms, and evaluation methods.

You'll collaborate with cross-functional ML teams, work with data annotation, and tackle new scientific problems with autonomy and impact across global offices, while remote-friendly policies support flexible work arrangements.