POSITION SUMMARY:

Natera's Real-World Data (RWD) ClinicalAI team turns clinical documents into structured data. The team has developed a platform that is now integral to Natera’s key initiatives leveraging RWD. In this role, you will own the experience of the people who work in the platform for their day to day operations. You will also own agentic workflows end to end: translating a stakeholder team's process into reliable, repeatable automation, deciding where AI agents act and where people interact, and making sure the final output is something the team can actually use.

This is an individual-contributor PM role: you'll shadow users, dig into the data yourself, write the requirements, and drive changes from idea to rollout, continuously focused on efficiency and innovation for your users. You will partner closely with engineering, software quality, and clinical teams that consume the data.

If you like making complex, high-volume workflows simple for expert users, turning AI agents into outputs people trust and act on, and seeing the improvement show up in the numbers, you will enjoy this role.

PRIMARY RESPONSIBILITIES:

Own the user-experience backlog and priorities for the user-facing application, including case review and editing, document navigation and evidence, work queues, and administrative review, in line with the overall platform roadmap.
Increase efficiencies without tradeoffs to accuracy: shadow end users, measure where the time goes, and validate problems before they become backlog commitments.
Turn intricate, multi-step review workflows into clear, intuitive interfaces.
Own how new AI capabilities appear to end users; define the experience, pilot with leads, and roll out without disrupting live production work.
Own agentic workflows end to end, from discovery through launch, validation, and iteration: translate a stakeholder team's needs into a phased workflow, decide where AI agents act and where people stay in the loop, and make sure the result is accurate, auditable, and efficient at scale.
Deliver usable outputs to stakeholder teams across the business: agree on what a finished output looks like and how success is measured, confirm it is used in their work, and close the loop when it is not.
Partner with the AI scientists and engineers to take agent prototypes to production: weigh evaluation results for user and stakeholder impact, decide how much autonomy a workflow is ready for, and own the pilot, rollout, and success metric.
Own the onboarding playbook for new user groups, including internal teams and vendor partners: access and roles, training cases in the test environment, team-lead setup, and support materials.
Plan user-facing releases so changes are tested, communicated, and adopted, with documentation and training built alongside clinical leads.
Replace manual work with self-service: own the user experience and rollout of an in-platform workflows and drive adoption of self-service project setup by delivery teams.
Define and track KPIs; use those metrics to guide priorities.
Own what happens after launch: measure adoption, iterate, and retire workarounds and manual processes once the product replaces them.

QUALIFICATIONS:

Bachelor's degree or equivalent practical experience, plus 5+ years of product management experience, including launching two or more user-facing software products or major releases.
Experience owning workflow software that expert users rely on all day, such as data abstraction or curation, annotation or labeling tools, clinical operations, or case management, where user productivity and data quality are the core metrics.
Strong, hands-on experience with agentic AI: you have defined and shipped products built on LLM-based agents or agentic workflows, and you understand evaluation, guardrails, and human-in-the-loop design well enough to make tradeoffs with engineers and AI scientists.
AI-native working style: you use AI tools every day to accelerate product discovery, synthesize user research, analyze data, draft requirements, prototype workflows, and pressure test ideas.