We are looking to hire a Prompt Validator based in India. This role requires a medically trained background — a nurse, doctor, or medical underwriter — to assist our prompt-writing consultant in validating the accuracy of AI-generated medical outputs and identifying areas for improvement.
Responsibilities
1. Clinical Data Auditing & Quality Assurance
• Check the AI Output: Review AI-generated medical summaries and structured data alongside the original, raw mediccal charts to ensure 100% accuracy.
• Catch Medical Errors: Identify missing diagnoses, incorrect medication dosages, confused timelines, or “hallucinations.”
• Underwriter Alignment: Ensure that the specific data points critical to insurance underwriters (e.g., severity of comorbidities, specific lab values, surgical history) are accurately captured and highlighted.
2. Prompt Refinement
• Error Analysis: When the AI makes a mistake, investigate why. Was the hospital note messy? Did the AI misunderstand a medical abbreviation?
• Prompt Writing: Work with our team to rewrite the AI’s instructions (prompts) in plain,
logical English to prevent the same mistakes from happening again.
• Continuous Improvement: Curate examples of highly complex medical records to continuously test the AI’s performance as it evolves.
3. Cross-Functional Collaboration
• Act as the bridge between the clinical/underwriting world and the software engineering team.
• Translate complex clinical concepts into straightforward rules that developers and machine learning models can follow.
Qualifications & Requirements
• Experience: 3+ years in clinical data abstraction, medical coding, clinical documentation improvement (CDI), nursing, or health informatics.
• Domain Expertise: Deep, native understanding of medical terminology, pharmacology, disease pathways, and how to navigate messy Electronic Health Records (EHR) or unstructured clinical notes.
• Meticulous Attention to Detail: A zero-tolerance policy for medical inaccuracies and a passion for