Craft, test, and refine prompts for Luna, the AI-powered digital health companion to deliver empathetic, clinically appropriate, and technically reliable responses.
Write, refine, and maintain prompts based on product specs, user personas, and clinical scenarios.
Use tools like OpenAI API, AWS Bedrock, and internal test frameworks to evaluate prompt performance and automate testing.
Rapidly iterate on prompts based on output analysis, user interaction data, and team feedback.
Collaborate with product managers, engineers, and clinical advisors to align tone, behavior, and therapeutic goals.
Conduct structured testing including A/B tests, hallucination checks, and performance scoring.
Contribute to internal prompt libraries, version control, and documentation best practices.
Help define and improve our prompt development methodology and safety checks.
Work closely with cross-functional teams to drive real-world impact for patients managing chronic conditions.
Requirements
Demonstrated success crafting high-quality prompts for LLMs, ideally with OpenAI’s API.
Strong grasp of prompt engineering workflows, prompt chaining, and prompt evaluation techniques.
Familiarity with prompt testing tools such as OpenAI Playground, AWS Bedrock, or LangChain.
Experience using tools like OpenAI API, AWS Bedrock, and internal test frameworks to evaluate prompt performance.
Analytical mindset with experience using structured testing, evaluation metrics, or model output analysis.
Excellent writing skills and attention to nuance, tone, and clarity in conversational UX.
Ability to conduct structured testing including A/B tests, hallucination checks, and performance scoring.
Experience with prompt libraries, version control, and documentation best practices.
Experience defining and improving prompt development methodology and safety checks.
Technical skills and familiarity with tooling to automate testing.
Ideally has built an LLM-based application with prompts that serviced millions of people.
Interest in healthcare technology and enthusiasm for improving patient outcomes through AI.
Comfortable working independently in a fast-moving, low-structure startup environment.