Director, Data Science - AI Product & Risk

About the role

The Director, Data Science - AI Product & Risk leads the strategic direction and operational excellence for AI across Enterprise Digital Product. They are responsible for setting the decision framework for risk, defining standards for Responsible AI governance, raising the bar on delivery, ensuring shipping with safeguards, owning monitoring & incident response, proving value with measurement, enabling teams to move fast, and leading and growing the team.

Responsibilities

• Set the decision framework for risk: Implement a tiered risk model that drives required reviews, test rigor, launch gates, and monitoring based on user impact and regulatory exposure
• Be a Product Leader of Responsible AI governance: Define standards for data use, documentation, evaluation, human oversight, accessibility, and customer transparency aligned with RAI, Legal, Privacy, Security, Compliance, and Brand
• Raise the bar on delivery (MLOps/LLMOps): Standardize the path from dev to prod, versioning, reproducibility, CI/CD, model/prompt registries, evaluation harnesses, and rollback strategies
• Ensure Shipping with safeguards: Partner in ensuring red-teaming, bias/fairness checks, privacy reviews, security testing (e.g., prompt injection), and guardrails for customer-facing experiences occur
• Own monitoring & incident response: Define telemetry, alerts, and SLOs; build AI runbooks for drift, data/pipeline failures, vendor outages, harmful outputs, and policy changes; lead post-incident reviews and corrective actions
• Prove value with measurement: Define KPIs, experiment design, and success metrics for every initiative tied directly to business outcomes
• Enable teams to move fast: Provide reusable patterns, templates, tooling guidance, and training so teams don't reinvent governance or delivery practices
• Lead and grow the team: Attract, mentor, and develop data scientists/ML engineers; set expectations and clear career pathways

Qualifications

• Proven experience leading data science/ML teams and delivering production AI in a product environment
• Solid understanding of modern ML and generative AI, including evaluation, monitoring, and lifecycle management
• Hands-on experience implementing Responsible AI practices (risk assessments, documentation, governance, audit readiness)
• Track record partnering with Product and Engineering to set strategy and deliver measurable outcomes
• Operational excellence: incident response leadership, reliability/SLO mindset, and crisp communication under pressure
• Ability to influence cross-functionally and drive alignment in ambiguous, fast-moving environments. The ability to articulate complex scientific concepts/language in a way that all stakeholders, partners, etc. can understand

Required Qualifications

• 5+ years hands-on experience with NLP and/or LLM data pipelines, training datasets, annotation frameworks, and data quality diagnostics specific to chat/voice systems; proven ability to trace data issues to model behavior impacts
• 5+ years proven track record of experience setting data strategies that support product teams and making trade-offs between safety investment and product velocity; translates regulatory/risk requirements into actionable roadmaps
• 3+ years leading response to or learned from a safety incident, bias discovery, or data breach; demonstrates systems thinking about root causes; shows genuine commitment to transparency and fixing underlying processes
• Fully authorized to work in the US without any restrictions
• Fluent in Security, Privacy, Legal, and AI Ethics; proven ability to bridge conflicting requirements across multiple risk functions and build durable governance models that multiple teams trust and enforce
• Proven ability to influence VPs and C-suite on complex decisions; comfortable presenting data/AI risk trade-offs to boards; demonstrates healthy disagreement and consensus-building across competing interests
• Led geographically distributed or matrixed teams across data, AI, and business functions; experience convening risk/policy stakeholders and making trade-off decisions that multiple functions can live with
• Successfully hired, developed, and led senior data and product professionals; provides technical direction while enabling autonomy; creates psychological safety for raising hard concerns
• Demonstrated track record building or scaling safety/governance frameworks for AI products; deep knowledge of bias, toxicity, hallucination, and jailbreak vectors; experience implementing measurable safety controls and audit mechanisms