Vice President, Data Science

Fidelity Investments•Boston, MA

About The Position

VP, Data Science – Quantitative Research, Measurement & Strategy Are you interested in leading the scientific backbone of a modern AI organization where rigor, measurement, and evidence drive strategy and execution? Fidelity Institutional is seeking a VP, Data Science to lead its Quantitative Research & Measurement function within the AI Center of Excellence (AI CoE). The VP of Data Science is accountable for how Fidelity Institutional measures impact, establishes truth, runs experiments, proves causality, and optimizes decisions at scale. This role requires deep, hands‑on proficiency with large language models, generative AI, and agentic systems, while ensuring their application is scientifically sound, empirically validated, and grounded in rigorous quantitative evidence. The VP of Data Science is accountable for ensuring insights derived from both traditional modeling and GenAI techniques are defensible, measurable, and decision‑relevant. This is a hands-on leadership role that sets the vision for advanced analytics as a center of excellence for measurement, experimentation, and quantitative decision science, partnering closely with Platform, Product, BI, Risk, and Business leaders.

Requirements

Master's or PhD in Statistics, Economics, Mathematics, Operations Research, Computer Science, or related quantitative discipline
15+ years of experience in advanced analytics, quantitative research, or data science
Proven leadership of senior quantitative teams and FI-level analytics programs
Deep expertise in statistics, probability, and experimental design
Strong background in causal inference and incremental impact measurement
Advanced knowledge of optimization, econometrics, and forecasting
Ability to assess modeling approaches for correctness, bias, and suitability
Advanced proficiency in Python for statistical modeling, experimentation, simulation, and analysis (NumPy, Pandas, SciPy, Statsmodels, Scikit-learn)
Strong working knowledge of SQL and large-scale analytical datasets (e.g., Snowflake)
Hands-on proficiency with large language models and generative AI, including prompt design, retrieval-augmented generation, structured outputs, agentic workflows, and quantitative evaluation of LLM behavior using task-specific metrics and statistical testing
Thinks like a scientist and leader: hypothesis-first, evidence-driven, and principled
Sets high bars for rigor, correctness, and interpretability
Comfortable challenging narratives that are not supported by data
Communicates complex modeling concepts to executive audiences with clarity
Creates alignment between quantitative truth and business action

Responsibilities

Define and own the vision for measurement, experimentation, and quantitative decision‑making
Establish standards for what should be measured and how impact should be proven across FI initiatives
Ensure consistent, defensible evaluation methodologies across analytics, AI, and business programs
Elevate data science from prediction to decision quality
Set strategy and standards for experimental design across the organization
Ensure statistical rigor in A/B testing, quasi‑experiments, and observational studies
Define best practices for power analysis, bias control, inference, and interpretation
Act as executive sponsor for experimentation platforms and methodologies
Own the FI approach to causal inference, attribution, and incrementality measurement
Ensure leaders can distinguish correlation from causation in decision‑making
Sponsor advanced causal techniques such as Difference‑in‑Differences, synthetic controls, and uplift modeling
Provide executive guidance on “Did it actually work?” questions
Establish optimization and decision‑science capabilities across FI
Guide formulation of objective functions, constraints, and trade‑offs aligned to business goals
Oversee deployment of optimization methods for prioritization, planning, and resource allocation
Ensure optimization outputs are interpretable and actionable
Set direction for hypothesis‑driven research to answer strategic business questions
Sponsor development of advanced statistical, econometric, and ML models where appropriate
Ensure models are theoretically sound, well‑documented, and fit‑for‑purpose
Promote scientific integrity and intellectual rigor across the AI CoE
Lead forecasting using time‑series and probabilistic techniques
Ensure uncertainty and scenario analysis are incorporated into forecasts
Partner with business leaders to integrate forecasts into planning and decision cycles
Lead recommendation approaches rooted in statistical learning, optimization, and behavioral science.
Ensure recommendation logic is explainable, empirically validated, and optimized against business and client outcomes rather than treated as black‑box ML.
Lead the design and evaluation of statistically grounded segmentation frameworks to uncover meaningful heterogeneity in clients, advisors, firms, and institutional behaviors.
Ensure segmentations are interpretable, stable, and actionable, with clear hypotheses for how segments drive differentiated strategy and outcomes.
Develop and govern probabilistic and causal models that estimate likelihood of action and incremental impact of interventions.
Own rigor around bias control, validation, and lift measurement to ensure models support decision‑making through incrementality.
Apply hypothesis‑driven analytics to understand longitudinal behavior, action sequences, and decision pathways across journeys, focusing on causal drivers and friction points.
Advance graph‑based analytics to model institutional, firm‑level, and advisor relationships with emphasis on influence, connectivity, exposure, and systemic effects supported by statistical validation.
Lead hands‑on design, experimentation, and evaluation of LLM‑ and agent‑based systems for knowledge extraction, classification, summarization, reasoning, and decision support.
Develop and implement task‑level evaluation frameworks, prompt and retrieval strategies, and controlled experiments to assess reliability, calibration, hallucination risk, bias, and robustness.
Build and test retrieval‑augmented generation and agentic workflows with explicit hypotheses about information quality and decision impact, and quantify incremental value relative to non‑generative statistical and machine‑learning baselines.
Build, lead, and mentor senior data science and quantitative research leaders who operate as scientific owners of modeling, inference, and measurement.
Define clear career paths and skill expectations for scientifically‑oriented data scientists, emphasizing statistics, causal inference, experimental design, decision science, and interpretability.
Foster a culture that values curiosity, peer review, principled debate, experimentation, and continuous learning in quantitative methods.
Serve as a trusted advisor to senior business and technology leaders
Translate complex quantitative findings into clear executive narratives
Influence strategy by grounding discussions in evidence, causality, and expected impact