Remote ML Safety & Evaluation Engineer

10a Labs; San Francisco (CA); On-site; USD 130,000 - 200,000; Full time

Benefits offered by this job

Performance-based annual bonus

Professional development support

Comprehensive health benefits

Fully remote, U.S.-based

Job summary

10a Labs is seeking a Machine Learning Engineer to design, build, and evaluate advanced ML systems for AI safety and model evaluation. You will work on reinforcement learning, NLP/LLMs, multimodal systems, and classifiers.

Collaborating with engineers, analysts, red teamers, and SMEs to turn ambiguous questions into rigorous experiments and scalable systems. This is a fully remote role based in the United States with a competitive salary and benefits.

Qualifications

• 3–5+ years in machine learning, research engineering, or related field.
• Strong Python skills with ML frameworks (PyTorch or JAX).
• Hands-on experience training, fine-tuning, or evaluating modern ML models.
• Solid understanding of experimental design, model evaluation, and quantitative analysis.
• Familiarity with agentic AI fundamentals, including Model Context Protocol and AI safety risks.
• Experience in reinforcement learning, NLP/LLMs, computer vision, or multimodal ML.

Responsibilities

• Design and run ML experiments to evaluate capabilities, behavior, robustness, and limitations of advanced AI systems.
• Develop and evaluate models across RL, NLP/LLMs, CV, and multimodal ML.
• Build evaluation pipelines, benchmarks, datasets, and metrics for frontier AI systems.
• Train, fine-tune, and evaluate models for safety, security, and other high-impact applications.
• Develop reliable tooling and infrastructure to run ML experiments and evaluations at scale.
• Analyze results, identify model failure modes, and translate findings into new experiments and approaches.

Skills

Python

PyTorch

JAX

ML experiments

Experimental design

Tools

PyTorch

JAX

Job description

10a Labs is seeking a Machine Learning Engineer to design, build, and evaluate advanced ML systems for AI safety and model evaluation. You will work on reinforcement learning, NLP/LLMs, multimodal systems, and classifiers.

Collaborating with engineers, analysts, red teamers, and SMEs to turn ambiguous questions into rigorous experiments and scalable systems. This is a fully remote role based in the United States with a competitive salary and benefits.