Verily
Staff Product Manager, AI for Care
$170K
San Bruno, California · Hybrid · Staff · Full-time · Posted Oct 26, 2026
- Experience
- 5+ years of product management on platform products.
- Visa sponsorship
- Qualified applicants must not require employer sponsored work authorization now or in the future for employment in the United States.
About the job
Owns the strategy and delivery of Verily’s AI Platform and Evaluation Harness, enabling developers to evaluate, monitor, and improve healthcare AI experiences.
What you'll do
- Set product vision, strategy, roadmap, and priorities for AI evaluation capabilities.
- Lead product discovery and delivery with engineering and AI teams, building scalable tools across customers and clinical risk levels.
- Align technical, clinical, regulatory, quality, security, and business stakeholders on platform decisions and compliance needs.
What they ask for
- 5+ years of product management on platform products, including embedded work with engineering teams and shipping and operating products.
- Hands-on experience shipping customer-used AI evaluation tooling, such as LLM-as-a-judge frameworks, golden datasets, code assertions, and trace tooling.
- Background in software engineering, systems engineering, data science, clinical informatics, or a comparable discipline.
- Experience aligning stakeholders and owning a multi-tenant, version-controlled platform with tenancy, versioning, and backward-compatibility tradeoffs.
Nice to have
- Healthcare domain fluency, including FHIR resources, clinical informatics, and interoperability standards.
- Experience with healthcare compliance frameworks or evaluation of clinical or high-stakes AI outputs.
Pay
USD $169,500-$254500 base salary
Tools
FHIR, HIPAA, GxP, FDA SaMD, LLM-as-a-judge
AI focus
Build and own AI evaluation tooling for healthcare, including golden-dataset evaluations, code assertions, LLM-as-a-judge evaluations, error analysis, customer feedback signals, and validation for regulated use.
Model-written summary · the original posting is authoritative