Principal Product Manager, Clinical

Role Overview: We are seeking a Principal Clinical Product Manager to serve as the clinical product lead for the Visión platform. You will own the translation of patient-level clinical reality — diagnoses, medications, laboratory results, utilization, and social factors — into safe, explainable, and actionable AI-driven predictive insights and workflows that care teams act on every day. This is a hands-on, senior individual contributor role that leads through influence. You will work at the intersection of clinical medicine, healthcare data, and data science: defining the computable phenotypes and cohort logic that determine who Visión surfaces, setting the reliability thresholds that determine what happens to a model output, designing the work queues and clinical summaries clinicians and care managers actually use, and holding the line on clinical safety and data quality as the platform scales across the Intelligence Loop — Identify, Recommend, Act, Monitor, Rescore.

Key Responsibilities: Own the product vision, roadmap, prioritization framework, and release strategy for Visión's clinical AI capabilities across ASCVD, COPD, CKD, and diabetes, and define product direction for new capabilities serving care delivery teams. Conduct discovery directly with clinical users — physicians, nurses, pharmacists, care and case managers, medical assistants, and population-health teams across primary, specialty, emergency, inpatient, pharmacy, and care management settings — to surface unmet needs and validate which capabilities earn a place in their day. Define the problem before defining the solution. Frame what is actually being solved and for whom, then translate customer, clinician, operational, and regulatory needs into clear requirements — user stories, workflow specifications, data definitions, acceptance criteria, and edge cases — and scope the minimum viable product without compromising clinical safety, data quality, or usability. Apply working knowledge of ICD-10-CM, CPT/HCPCS, NDC/RxNorm, LOINC, SNOMED CT, and DRG/HCC risk adjustment to define cohorts and surface meaningful patterns — distinguishing active from historical diagnoses, confirmed from rule-out, chronic disease from encounter diagnosis, and a single abnormal result from a clinically meaningful trend. Define each AI model's intended use, target population, prediction horizon, outcome, and associated intervention, and partner with data science, clinical, and engineering to specify features, labeling strategy, training and validation data, ground truth, performance requirements, and monitoring and release criteria. Establish measurable outcomes for every feature — better identification of eligible patients, reduced chart-review time, increased care-gap closure, improved post-hospitalization follow-up, reduced avoidable utilization, higher clinician adoption, improved model precision. Translate clinical concepts into computable phenotypes, business rules, data requirements, and acceptance criteria — including inclusion/exclusion criteria for chronic-disease populations built from diagnosis, procedure, medication, laboratory, encounter, and utilization data. Lead backlog refinement and sprint prioritization, translating clinical research into clear direction for data science and engineering on what to build next. Own clinical data-quality expectations for the platform and define clinical data exchange requirements. Determine when a model output is reliable for clinician review, work queues, patient outreach, population reporting, or automated action; design human-in-the-loop escalation for uncertain or high-risk outputs; and ensure scores and recommendations are explainable in practical clinical terms. Apply Claude Code and similar AI-assisted development tools to prototype concepts and accelerate design iteration. Explain complex clinical and AI concepts credibly to executives, clinicians, customers, engineers, and data scientists; support executive briefings, sales presentations, and strategic planning with senior payer and provider leadership; and prepare the organization to bring capabilities to market through demonstrations, enablement and marketing materials, product documentation, workflow diagrams, data dictionaries, release notes, training, and implementation guidance for commercial and government customers.

Qualifications: Bachelor's degree and 7+ years in healthcare analytics, clinical informatics, population health, or healthcare technology product management, with substantial hands-on work in patient-level clinical data. Demonstrated expertise in ICD-10-CM and related code sets, computable phenotypes and cohort development, data-quality assessment, and chronic-disease workflows, plus experience partnering with data science and engineering to develop, validate, and deploy AI or predictive-analytics products in clinical or operational settings. Extensive, hands-on knowledge of patient-level clinical EHR and claims data and its use in analytical and predictive products — able to query and profile the data directly, recognize the patterns and relationships within it, judge what it can and cannot support, and translate that understanding into new solutions. This role requires deep fluency in the data itself; general healthcare industry experience is not a substitute. Experience conducting workflow discovery and validation directly with clinicians and care teams — physicians, nurses, pharmacists, care managers, population-health staff — and designing patient lists, work queues, registries, or alerts that fit inside their existing day, with demonstrated attention to unsafe failure modes and alert fatigue. Demonstrated experience building and validating patient cohorts and computable phenotypes from diagnosis, procedure, medication, laboratory, encounter, and utilization data, with defensible inclusion and exclusion criteria. Advanced working knowledge of ICD-10-CM, CPT/HCPCS, NDC/RxNorm, LOINC, SNOMED CT, and DRG/HCC risk adjustment, with the judgment to distinguish active from historical diagnoses, confirmed from rule-out, and a single abnormal result from a meaningful trend. Proven ability to write the artifacts the team builds from — user stories, workflow specifications, data definitions, acceptance criteria, edge cases — and to lead backlog refinement and sprint prioritization in an Agile environment. Hands-on experience with Claude Code, Cursor, or similar AI-assisted development tools for rapid prototyping and design iteration. Direct experience taking a model from problem statement to production with data scientists and ML engineers: intended use, target population, prediction horizon, outcome, labels, ground truth, and release criteria. Track record diagnosing and mitigating real-world data problems — duplicates, incomplete or delayed documentation, coding artifacts and inconsistency across organizations, terminology version changes, upcoding risk — and specifying product behavior when clinical evidence conflicts. Preferred experience with at least two of ASCVD, COPD, CKD, and diabetes., including how comorbidities, medications, labs, utilization, and social factors shape risk and progression. Experience taking an AI-enabled healthcare product through customer implementation in a regulated SaaS, health-tech, payer, provider, or government environment (CMS, VA, DHA a plus).Ability to evaluate model output using both clinical and statistical reasoning — setting the threshold at which a prediction is trustworthy, determining where human review belongs, and making output explainable to clinicians in terms of underlying clinical evidence rather than feature weights. Experience partnering with compliance, privacy, security, and regulatory functions under HIPAA, minimum-necessary use, and role-based access. Exceptional written and verbal communication, including product demonstrations and briefings to clinical and executive audiences at customer organizations. Preferred working knowledge in the areas of population health, value-based care, ACO, payer analytics, case management, or quality improvement background, including CMS quality measures, HEDIS, MIPS, Stars, risk adjustment, or utilization management. Preferred experience integrating clinical data from Epic, Oracle Health/Cerner, or payer-provider platforms. Preferred experience evaluating generative AI or NLP for record summarization, coding abstraction, or care-gap detection.