Design and review AI training tasks based on life sciences research, experiments, datasets, and evidence. Work remotely in English with a schedule requiring at least 40 hours each week and PST overlap.
The Work
You will help train AI systems to reason about life sciences. The work uses research scenarios, experiments, datasets, and scientific evidence to create structured examples and clear expert feedback.
You will write or improve responses, review source materials and model outputs, and explain your judgments so the feedback can improve model behavior. You will also maintain consistent quality across repeated review tasks.
- Write or improve responses to life sciences scientific reasoning tasks.
- Review source materials and AI-generated outputs against project guidelines.
- Explain your reasoning and judgments clearly in writing.
- Apply rubrics consistently across repeated evaluation tasks.
What It Pays and Takes
Compensation is handled through the OpenTrain project budget fields shown on the job page. The role is a remote contractor position, and prior AI training experience is not required.
- Work type: Contractor and part-time.
- Schedule stated for the role: at least 8 hours per day and at least 40 hours per week.
- The listing also specifies a minimum time requirement of 20 hours per week.
- Time zone: 4 hours of overlap with Pacific Standard Time.
- Location: Worldwide.
- Language: Fluent English.
- Bring professional, academic, or language expertise in life sciences scientific reasoning.
- Strong written communication and close attention to detail are required.
- You must be able to follow detailed instructions, apply rubrics consistently, and work independently on remote tasks.
How It Works
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI Training Work
AI training is the human work behind systems that learn from examples, including writing and rating model responses. People with subject expertise are paid to provide careful judgments that help AI handle specialized questions more accurately.