Machine Learning Engineer for AI Model Evaluation

Help evaluate and improve frontier AI coding models by completing structured technical assessments that simulate realistic machine learning engineering workflows. This role focuses on using and critiquing coding agents to surface bugs, failure modes, and deployment tradeoffs, with limited spots available on a first come, first serve basis.

• Use frontier AI coding agents to complete and evaluate complex ML and AI engineering tasks.
• Review model-generated implementations related to model training, inference systems, MLOps, and LLM applications.
• Identify bugs, edge cases, performance problems, and failure modes in model outputs and implementations.
• Compare outputs from multiple frontier models, assessing their relative strengths and weaknesses.
• Apply professional engineering judgment to realistic ML engineering scenarios and recommend actionable improvements.

• At least 2 years of professional machine learning engineering experience.
• Experience building production ML systems, model deployment infrastructure, LLM applications, or AI-powered products.
• Regular use of AI coding agents such as Cursor, Claude Code, Codex, Windsurf, Gemini CLI, or similar tools.
• Ability to evaluate model-generated ML implementations and reason about technical tradeoffs.
• Experience deploying ML systems to production is preferred.

• Remote engagement.
• Employment type, hourly.
• Sprint-based project work, with assignments running in 12 to 24 hour stretches depending on client requirements.
• Spots are limited and are filled on a first come, first serve basis.
• Task availability and placement depend on client needs.

• Listed hourly rate: $85 per hour.
• Alternate payment detail: $400 per accepted task, with typical tasks taking approximately 2 to 3 hours after ramp-up.
• Compensation is tied to accepted work.

• Ideal candidates have 2 or more years of professional ML engineering experience and relevant production deployment experience.
• Regular practical experience with AI coding agents and with evaluating model-generated engineering outputs is required.