The role
From Gilead Sciences's own posting.
At Gilead, we’re creating a healthier world for all people. For more than 35 years, we’ve tackled diseases such as HIV, viral hepatitis, COVID-19 and cancer – working relentlessly to develop therapies that help improve lives and to ensure access to these therapies across the globe. We continue to fight against the world’s biggest health challenges, and our mission requires collaboration, determination and a relentless drive to make a difference.
Every member of Gilead’s team plays a critical role in the discovery and development of life-changing scientific innovations. Our employees are our greatest asset as we work to achieve our bold ambitions, and we’re looking for the next wave of passionate and ambitious people ready to make a direct impact.
We believe every employee deserves a great leader. People Leaders are the cornerstone to the employee experience at Gilead and Kite. As a people leader now or in the future, you are the key driver in evolving our culture and creating an environment where every employee feels included, developed and empowered to fulfil their aspirations. Join Gilead and help create possible, together.
Job Description
At Gilead our pursuit of a healthier world for all people has yielded a cure for hepatitis C, revolutionary improvements in HIV treatment and prevention as well as advancements in therapies for viral and inflammatory diseases and certain cancers.
We set and achieve bold ambitions in our fight against the world’s most devastating diseases, united in our commitment to confronting the largest public health challenges of our day and improving the lives of patients for generations to come. As a Director, AI Systems Quality & Governance, R&D and Gilead, you will establish and operationalize the quality framework for artificial intelligence across Research & Development.
Reporting to the Senior Director of eSystems Quality for R&D, this role will define how AI-enabled solutions are assessed, developed, tested, implemented, monitored, changed, and retired within a regulated environment. The Director will ensure that quality, data integrity, risk management, and appropriate controls are built into AI solutions by design. This individual will also establish the documentation, testing, traceability, and ongoing monitoring needed to demonstrate that AI solutions remain fit for intended use and inspection ready throughout their lifecycle.
Extensive experience within a Quality organization is not required. We are seeking someone with strong experience in AI, data, digital products, technology risk, or controls who is willing to develop a deeper understanding of GxP regulations and apply guidance from colleagues with expertise in Quality, validation, data integrity, and regulated R&D processes.
This is an onsite position that operates within our G.Flex policy (hybrid, on site Tuesday-Thursday).
Overview of Responsibilities:
Establish the AI Development Lifecycle
Design, implement, and maintain a risk-based AI Development Lifecycle for R&D.
Define requirements for intake, intended use, risk classification, data readiness, development, testing, release, monitoring, change management, periodic review, and retirement.
Establish required documentation, evidence, reviews, approvals, and decision points across the lifecycle.
Ensure that quality and control requirements are incorporated during solution design and development rather than added at the end.
Align AI lifecycle requirements with applicable GxP regulations, Gilead procedures, data-integrity principles, and inspection expectations.
Define AI Risk and Control Requirements
Develop a risk and control framework for AI-enabled solutions.
Evaluate risks associated with model performance, unreliable outputs, bias, data quality, drift, explainability, automation, human oversight, privacy, security, and vendor-controlled changes.
Define controls that can be implemented, tested, monitored, and supported by objective evidence.
Establish requirements for human review, intervention, override, escalation, and documentation.
Determine how control effectiveness will be demonstrated throughout the lifecycle.
Work with GxP Quality subject-matter experts to determine when formal change control, deviation, investigation, CAPA, or other Quality-system processes apply.
Establish Data Quality Requirements
Define requirements for data used to train, configure, test, validate, and operate AI solutions.
Develop a framework for grading data based on accuracy, completeness, consistency, representativeness, lineage, provenance, timeliness, criticality, and fitness for intended use.
Ensure that data sources, transformations, exclusions, assumptions, and limitations are documented.
Define traceability expectations across source data, model versions, requirements, testing, outputs, and production use.
Evaluate how patient, clinical, proprietary, and company data may be used by internal models, external platforms, and vendors.
Define AI Testing and Assurance Strategies
Establish risk-based testing and assurance approaches for AI-enabled solutions.
Define expectations for requirements, acceptance criteria, test data, test methods, traceability, and objective evidence.
Ensure that test strategies address accuracy, reliability, reproducibility, boundary conditions, failure scenarios, bias, human oversight, and real-world performance.
Define how teams will demonstrate that AI-generated outputs are correct, appropriate, and reliable for their intended use.
Determine when statistical testing, expert review, comparative testing, deterministic testing, or production monitoring is appropriate.
Ensure that model limitations, prohibited uses, expected failure modes, and required user actions are documented.
Establish Monitoring and Periodic Review
Define requirements for monitoring AI solutions after implementation.
Establish expectations for monitoring model performance, model drift, data drift, output quality, exceptions, human overrides, unintended outcomes, and vendor changes.
Define meaningful metrics, thresholds, alert criteria, and escalation pathways.
Establish triggers for investigation, retraining, additional testing, revalidation, suspension, or retirement.
Define risk-based periodic review requirements focused on continued fitness for intended use.
Ensure that monitoring responsibilities and decision rights are clearly assigned.
Evaluate AI Models, Platforms, and Vendors
Assess the risks and controls associated with open, closed, hosted, embedded, and vendor-managed AI models.
Evaluate vendor documentation related to model design, training data, performance, security, privacy, monitoring, and change management.
Determine whether vendor evidence is sufficient to support Gilead’s intended use.
Identify additional testing, controls, contractual provisions, or monitoring needed to address gaps.
Define expectations for vendor notification of model updates, retraining, material changes, incidents, and performance issues.
Evaluate whether company data or user inputs may be retained or used to train external models.
Ensure Inspection Readiness
Establish documentation and traceability expectations for AI-enabled solutions.
Ensure that intended use, risks, requirements, testing, approvals, limitations, monitoring, changes, and decisions are supported by retrievable evidence.
Confirm that assurance activities are proportionate to risk and defensible during internal audits and health authority inspections.
Help technical and business teams explain how AI solutions work, how risks are controlled, and how continued performance is demonstrated.
Identify and address documentation, testing, or control gaps before inspection.
Apply lessons learned from audits, inspections, incidents, and implementation experience to improve the AI Development Lifecycle.
Build AI Quality Capabilities
Increase understanding of AI technologies, risks, and controls across R&D and eSystems Quality.
Develop playbooks, templates, decision tools,