Position Summary

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. As MIDAS grows and AI workflows are introduced, this person will ideally help build out that program under the direction of BI and prepare it for handoff and training to Operations.

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 simple rules that developers and machine learning models can follow.

Requirements

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 spotting inconsistencies that others miss.
• Tech-Forward Mindset: Highly computer literate, adaptable to new software, and eager to learn how Large Language Models (LLMs) like ChatGPT work behind the scenes.
• Communication: Exceptional written communication skills — able to explain complex medical concepts simply and write clear, unambiguous instructions.

Preferred

• Previous experience in quality assurance (QA) or auditing roles.
• Familiarity with life or health insurance underwriting processes and risk factors.
• Personal or professional experimentation with AI tools (ChatGPT, Claude, etc.) to automate daily tasks.