Prompt Testing Assistant

About the role

Braun Management is seeking a part‑time, fully remote Prompt Testing Assistant to help ensure that our AI‑driven research tools generate accurate, reliable outputs. You will work closely with senior consultants to validate prompt performance against internal datasets and maintain high standards of factual precision.

Key responsibilities

• Test specific prompt variations against internal datasets to identify inconsistencies and logic errors.
• Document the quality and factual accuracy of text‑based summaries and data categorizations.
• Compare different model versions to determine which handles complex governance and business terminology most effectively.
• Maintain a structured log of prompt failures and unexpected behaviours for the technical support team.
• Review generated text for appropriate tone and clarity according to established writing style.
• Assist in developing standardized templates for repetitive analytical research tasks.
• Verify that model outputs adhere to strict constraints and do not include fabricated figures or events.

Required profile

• Strong interest in how language models can be applied to business and organizational research.
• High attention to detail when reviewing long texts for small inconsistencies.
• Academic or project experience that demands rigorous reading and critical analysis.
• Ability to work independently in a fully remote environment with clear written instructions.
• Willingness to learn and apply prompt‑engineering concepts.
• Clear communication skills to explain why a specific output meets or fails to meet standards.

Required skills

• Basic familiarity with prompt engineering concepts.