AI Behavior Trainer

Role overview Help improve how AI systems respond by reviewing their outputs and applying ethical standards across different scenarios. The work combines careful evaluation, training data creation, and collaboration with engineering and quality teams to strengthen model safety and reliability.

Responsibilities - Assess and revise model responses for quality and responsible behavior. - Identify risks such as bias, unsafe content, and misleading answers. - Create training examples and develop prompts and evaluation criteria. - Review output patterns to find weaknesses and emerging risks. - Work with machine learning and quality teams to refine training processes. - Provide iterative feedback to improve model alignment and reliability.

Requirements - Strong critical thinking, attention to detail, and written communication. - Ability to assess language, reasoning, and context accurately. - Understanding of fairness, safety, and responsible technology use. - Consistency in applying detailed guidelines and evaluation frameworks. - Interest in how AI systems behave in real-world settings.

Nice to have - Experience in data annotation, AI training, or content evaluation. - Familiarity with AI systems, large language models, model alignment, prompt design, or AI safety. - A degree in communications, psychology, linguistics, philosophy, or education.

Benefits and work setup - Remote independent contract work, four days per week, on a shifting schedule. - Structured work cycles focused on experimentation and development.