ML Systems Engineer - Prompting & Debugging ML Deployments

DataAnnotation; Portsmouth (NH); Remote; USD 170,000 - 243,000; Full time

Benefits offered by this job

Flexible, remote, and project-eligible

Job summary

DataAnnotation is seeking a Machine Learning Engineer to stress-test how models reason about ML systems, including training, evaluation, data pipelines, and deployment trade-offs. You’ll craft prompts and critique model outputs, guiding how practitioners should approach similar challenges.

In this role you’ll write correct solutions grounded in real experience, review AI results for subtle errors, and contribute to robust, transparent ML tooling. Flexible, remote project work is available.

Qualifications

• Hands-on experience training, evaluating, or deploying models professionally or in serious personal work.
• Comfort with the modern ML stack; most tasks assume Python and PyTorch or JAX.
• Clear written English: your explanations are the training signal.
• No degree required. We care about what you can do, not where you learned it.

Responsibilities

• Write prompts that probe how models reason about training, evaluation, debugging, and productionizing ML systems.
• Review AI output for subtle errors: leaky evaluations, wrong loss formulations, and misdiagnosed training failures.
• Write the correct solution when the model falls short, grounded in real practitioner experience.

Skills

Python

Clear written English

ML systems

Hands-on ML experience

Prompt engineering

Review ML outputs

Tools

PyTorch

JAX

Python

Job description

DataAnnotation is seeking a Machine Learning Engineer to stress-test how models reason about ML systems, including training, evaluation, data pipelines, and deployment trade-offs. You’ll craft prompts and critique model outputs, guiding how practitioners should approach similar challenges.

In this role you’ll write correct solutions grounded in real experience, review AI results for subtle errors, and contribute to robust, transparent ML tooling. Flexible, remote project work is available.