Vice President, AI Research and Real-World Evidence

Flagship Pioneering, Inc.•Cambridge, MA

3d•$263,000 - $346,500

About The Position

FL113 is a fast-moving AI diagnostics company with a bold ambition: to define the future of pre-emptive and proactive care. We are building products that help uncover disease before it becomes visible, equip patients and care teams with earlier insight, and enable meaningful intervention before disease advances and causes irreversible harm. Our team brings together ML researchers, serial entrepreneurs, and clinical experts, united by a common commitment to make proactive care a reality. FL113 is part of the Flagship Pioneering ecosystem, a premier venture-creation platform recognized twice on FORTUNE’s “Change the World” list and twice on Fast Company’s list of the World’s Most Innovative Companies. Flagship creates companies at the frontier of science and technology, bringing together founders, operators, and investors to take bold leaps on consequential problems and turn breakthrough ideas into enduring businesses. Join an exceptional team and help shape the future of AI and healthcare. The Vice President of AI Research and Real-World Evidence owns FL113’s scientific agenda, model methodology and performance, clinical validation, and evidence generation. You will determine which scientific questions to pursue, how to assess model performance and clinical utility, and what evidence is required to support product readiness, customer adoption, and regulatory strategy. You will work closely with peer leaders across Engineering, Product, and GTM, as well as customers, health-system partners, and the broader Flagship ecosystem, to shape product development and delivery. You will retain accountability for scientific rigor, model performance, and real-world evidence. Research defines scientific requirements and acceptance criteria; Engineering owns production architecture, implementation, deployment, and operations. Together, we will use evidence from models, pilots, and customers to refine product-market fit and build products that earn trust in the real world.

Requirements

PhD or master’s degree in biostatistics, computational biology, epidemiology, mathematics, computer science, or another relevant quantitative discipline, combined with deep healthcare or life-sciences experience.
At least 12 years of experience across AI research, health data science, clinical evidence generation, or product development, including success in early-stage or zero-to-one environments.
Deep hands-on expertise in machine learning for structured and unstructured health data, including causal inference, deep learning, survival analysis, and longitudinal modeling.
Experience taking clinical AI models from research through real-world validation and into production in partnership with engineering teams.
Experience designing, executing, and interpreting retrospective and prospective clinical studies.
Fluency with modern research and machine learning tools, including Python, PyTorch, scikit-learn, and AI-enabled development workflows.
Strong working knowledge of health data sources, including electronic health records, claims, omics, wearables, and digital biomarkers.
Working knowledge of healthcare data standards and clinical vocabularies such as FHIR, HL7, OMOP, SNOMED, LOINC, ICD, CPT, and RxNorm.
Familiarity with healthcare privacy, security, and compliance requirements, including HIPAA, HITRUST, and SOC 2.
Experience communicating with clinicians, key opinion leaders, customers, executives, and other technical and nontechnical audiences.

Nice To Haves

You thrive in ambiguity and are energized by zero-to-one company building.
You pair scientific rigor with sound entrepreneurial judgment, comfortable making consequential decisions amid the uncertainty and imperfection of an early-stage venture.
You have a track record of building, leading and inspiring high-performing, multidisciplinary teams in fast-moving environments.
You communicate complex methods, results, limitations, and decisions with clarity and relevance across audiences – whether they are research scientists or business executives.
You are AI-native, relentlessly applying agentic tools and workflows to accelerate research, sharpen experimentation, and amplify your team’s output.

Responsibilities

Set the Scientific Agenda: Own and continuously evolve a multi-year research roadmap spanning predictive modeling, causal inference, and longitudinal analysis, informed by scientific, clinical, product, and market evidence.
Translate the research roadmap into practical milestones, decision gates, and resource requirements.
Exercise pragmatic scientific and product judgment, connecting model performance to clinical value, customer ROI, and development time.
Define and uphold fit-for-purpose standards for data quality, model validation, study design, reproducibility, and responsible use of AI.
Apply AI-enabled tools and agentic workflows to accelerate research, experimentation, analysis, and team productivity.
Lead Model Development and Validation: Define the scope and design of research programs, including indication selection, data strategy, modeling approach, endpoints, and validation criteria.
Oversee model development and validation across structured and unstructured health data, including causal inference, deep learning, survival analysis, and longitudinal modeling.
Own model methodology, evaluation plans, research code and prototypes, and the evidence required to establish performance and clinical utility.
Review model results and limitations, mentor research scientists, and uphold scientific quality across the team.
Define research dataset requirements, analytical schemas, feature definitions, and data-quality criteria in partnership with Engineering.
Establish Product Readiness: Define scientific acceptance criteria and productization requirements with Product and Engineering.
Develop model-readiness and validation plans that specify what must be demonstrated before a model moves into production or a customer pilot.
Provide scientific requirements for interoperability, data architecture, deployment, and monitoring while Engineering retains ownership of the production stack.
Confirm that production implementations preserve validated model behavior and that monitoring plans can detect clinically meaningful performance changes.
Inform feature requirements and launch readiness using model evidence, clinical utility, and customer needs.
Build Clinical Evidence and Scientific Credibility: Design and oversee retrospective and prospective clinical studies with health-system partners.
Establish the real-world evidence strategy for evaluating product performance, clinical utility, workflow impact, and economic benefit.
Shape the scientific design and success criteria for pilots with biopharma companies, accountable care organizations, health systems, and other early customers.
Partner with clinical and regulatory experts to develop evidence plans supporting clinical decision support and software-as-a-medical-device pathways.
Work with GTM to define the scientific scope, data requirements, and evidence commitments for customer and research collaborations.
Communicate results, limitations, scientific risks, and technical opportunities clearly to internal and external stakeholders.
Build FL113’s scientific credibility through engagement with key opinion leaders, publications, presentations, and external collaborations when appropriate.
Build and Lead the Research Organization: Recruit, lead, and develop a multidisciplinary team spanning machine learning, clinical research, real-world evidence, biostatistics, and health economics.
Set research priorities, direct day-to-day scientific operations, review technical work, and remove obstacles.
Create a culture that combines scientific rigor with the speed and practical judgment required in an early-stage company.
Establish effective working practices across Research, Engineering, Product, Clinical, Regulatory, and GTM.
Develop team members through direct technical mentorship, clear expectations, and accountability for reproducible results.