Strategic Projects Lead, Safety

About the role

Handshake is seeking a Strategic Projects Lead to manage large-scale human data programs that drive the training and evaluation of frontier AI models. You will design operational workflows for model policy annotation and influence safety enforcement decisions for leading AI systems. This role requires an operator who can navigate ambiguity and deliver results in a fast-paced environment. You will translate complex safety requirements into scalable execution plans that balance precision with speed. The position demands comfort with unstructured challenges and the ability to synthesize feedback from diverse stakeholders. You will be a central point of coordination for high-stakes safety initiatives that define model behavior. Success in this role requires a proactive mindset and the capacity to drive initiatives from concept to implementation.

Key facts

What you'll do

• Architect annotation frameworks and labeling workflows that prioritize high precision and inter-rater agreement at scale.
• Manage and mentor a team of annotators, ensuring consistency and accuracy while navigating complex edge cases and sensitive policy topics.
• Work with policy leads to provide actionable insights regarding classification ambiguities and gaps identified during data annotation.
• Develop and refine annotation playbooks and auditor resources to ensure alignment across regulatory domains and datasets.
• Identify systemic labeling errors and emerging data trends to improve internal safety policies.
• Engage directly with senior stakeholders and frontier AI labs to define project scope and resolve operational issues.
• Design and implement quality assurance mechanisms that detect anomalies in annotation outputs before they propagate.
• Facilitate cross-functional alignment sessions to reconcile competing priorities between safety, product, and research teams.
• Conduct regular reviews of operational metrics to assess the health and reliability of annotation pipelines.
• Lead incident response efforts when critical safety issues are identified in labeled data or model outputs.
• Partner with engineering teams to integrate annotation workflows into automated tooling and monitoring systems.
• Forecast resource needs and staffing requirements for upcoming safety evaluation cycles.
• Document standard operating procedures to institutionalize best practices and reduce dependency on individual expertise.
• Provide training and guidance to new team members on policy standards and annotation protocols.
• Contribute to external-facing materials that communicate safety processes to regulators and industry partners.

Requirements

• 2+ years of experience in trust and safety, policy enforcement, regulatory compliance, risk management, management consulting, or government.
• Analytical problem-solving skills with the ability to use SQL or similar tools to extract insights from large datasets.
• Strong communication skills with experience managing relationships with senior stakeholders.
• High level of personal accountability and an ownership-driven mindset.
• Demonstrated ability to work on-site in Seattle, Washington, with regular availability during standard business hours.
• Willingness to engage with sensitive, explicit, or disturbing material as part of safety and policy evaluation duties.
• Compliance with company policies regarding confidentiality and data handling in security-sensitive environments.
• Capacity to commit to the full duration of project cycles without interruption, given the operational nature of the work.
• Adherence to ethical guidelines when handling human-generated content and model training data.
• Understanding of the implications of model outputs on real-world safety and societal impact.
• Willingness to follow structured processes while contributing to iterative improvements of workflows.
• Eligibility to receive compensation in USD and participate in equity compensation plans as defined by company policy.
• Access to a reliable internet connection and suitable workspace to support full-time on-site responsibilities.

Nice to have

• Experience scaling content review or policy enforcement workflows across large, distributed teams.
• Subject matter expertise regarding high-severity policy harms, including sexual activities, violent acts, or dangerous organizations.
• Experience building evaluations for policy enforcement or harmful content use-cases.

Practical notes

This position involves regular exposure to sensitive, explicit, or disturbing material as part of the safety and policy evaluation process. Benefits for US employees include 401(k) matching, equity, medical, dental, and vision insurance, mental health support, a $500 wellness stipend, a $2,000 learning stipend, flexible PTO, 15 holidays plus 2 flex days, paid parental leave, and fertility benefits. The role requires full-time on-site presence in Seattle with regular hours and limited travel expectations. Candidates must be authorized to work in the United States without sponsorship for this position.