Product Operations, Pomelo Care

• Demonstrated interest or experience in AI/LLMs (e.g., prompt engineering, RAG, embeddings, knowledge graphs, vector databases)
• Strong information organization skills: detail-oriented, logical thinker, able to design systems for categorizing and structuring knowledge
• Technical literacy — comfortable working with structured content (Markdown, JSON, SQL basics, or similar)
• Curiosity about healthcare and clinical workflows (you do not need to be a clinician or but must be eager to learn)
• Strong communication skills, able to bridge technical and clinical conversations

Desirable

• Your North Star: help design, structure, and maintain the clinical, operational, and product knowledge that powers both our clinicians and our AI tools
• Designing & Structuring Knowledge
• Organize clinical guidelines, operational workflows, and product documentation into structured, AI- and clinician-friendly formats
• Create taxonomies, tagging systems, and metadata to make knowledge searchable and modular by both clinicians and AI agents
• Making Knowledge AI-Ready
• Partner with product and engineering to prepare knowledge for use in LLM-powered tools (e.g., use of Vertex AI for data organization, RAG pipelines, embeddings, structured content)
• Re-architect existing content into smaller, structured, referenceable units
• Maintaining & Governing Knowledge
• Ensure accuracy, version control, and consistent formatting across knowledge sources
• Audit and update content regularly in partnership with clinical and operational stakeholders
• Bridging Clinical and Technical Teams
• Translate clinical workflows and guidelines into structured knowledge for technical teams
• Provide feedback to engineers and data scientists on how knowledge is being consumed by AI tools
• Coordinate closely with Clinical Education team who manages clinician-facing knowledge including onboarding to ensure alignment of knowledge
• Support clinicians who are using AI tools and accessing organizational knowledge by managing inbound requests/feedback from #help-product