Customer Support Reps - AI Training - Oklahoma City, US

Role overview Use customer service experience to help train and assess AI systems through paid, task-based work. The work focuses on judging the quality of automated support and creating realistic examples that test how a system responds to customer needs. Applicants complete a short skills assessment before being invited to participate.

Responsibilities - Assess AI-written support replies for accuracy, clarity, helpfulness, and appropriate tone. - Create sample tickets and conversations that test responses to difficult or unusual situations. - Check generated help articles and reusable replies for correctness and ease of use. - Label responses for empathy, professionalism, and other aspects of customer sentiment. - Compare outputs with established support quality standards and identify errors.

Requirements - Professional experience in high-volume customer support, technical support, or client success. - Strong written English, with the ability to adjust tone and communicate clearly with different customer types. - Experience calming difficult interactions and explaining technical or billing troubleshooting step by step. - Fast, accurate typing and the ability to find relevant information across multiple sources. - Close attention to factual accuracy, reasoning, and tone in written responses. - A PayPal account to receive payments.

Nice to have - Fluency in additional languages, especially for assessing translations and localized support. - Familiarity with customer service platforms such as Zendesk, Salesforce Service Cloud, Intercom, or Freshdesk; collaboration tools; and knowledge bases such as Confluence, Notion, or Guru.

Benefits and work setup Tasks are paid, with rates varying by project; the source says some researchers offer up to $30 per hour. Be prepared for assignments requiring an uninterrupted hour, although some are shorter. The work is described as flexible and remote, with task availability dependent on researchers' needs.