Work location: on-site, Paris (France)

Engagement model: Temporary Employment (3 months, with possible extension)

Workload: 25 hours/week

Project start: July 2026

Languages required: Swedish and English

DataForce by TransPerfect is seeking a detail-oriented and linguistically skilled professional to design and evaluate user prompts for AI language models. The primary focus of this role is to create high-quality prompts, define grading criteria, and review model outputs to ensure accuracy, relevance, and adherence to guidelines. You will critically assess model generations, identify failure cases, and provide clear, well-justified feedback to improve system performance.

Tasks

• Develop user prompts and establish robust evaluation criteria for AI model testing.


• Review and rate model responses, ensuring complianc...