AI training jobs for machine learning engineers

These are the roles where your own field is the subject. Postings cover machine learning and ML engineering, AI engineering, data science and MCP work, and they pay at the top of the software range. The work is a mix: evaluating model output against a rubric, writing prompts and refining them, judging whether an agent's trajectory was sound, and ordinary engineering: bug fixing, refactoring, feature work, on codebases used to generate training data. Listed skills include machine learning, Python, pandas, statistics, RLHF, prompt engineering and AI model evaluation, alongside the same Java, Rust and TypeScript that appear on the general software roles. What gets judged is whether your evaluation is reproducible: another engineer applying your reasoning should land on the same score. Remote, hourly or per task, contract. Each posting lists its own rate.