Project Manager – Gen AI Frontier Lab

About the job

Project Manager – Gen AI Frontier Lab Strategic Delivery Role | Frontier Model Clients | Research-Driven AI Programs Role Snapshot This role leads high-ambiguity, research-intensive Gen AI data programs for frontier AI labs. The Project Manager operates at the intersection of research intent and scalable execution, driving structured experimentation, quality governance, and operational excellence. What Makes This Role Critical for Frontier Clients Operates in experiment cycles, not fixed-scope delivery models. Translates evolving research hypotheses into scalable execution workflows. Drives signal quality over raw volume. Partners directly with Research PMs, Safety Teams, and Evaluation Leads. Maintains delivery control amid rapid changes to rubrics and objectives. Core Responsibilities 1. Experiment-Driven Delivery Ownership Convert research objectives into sprint-based execution plans. Manage SFT, RLHF, preference ranking, safety tuning, and evaluation datasets. Maintain experiment logs capturing intent, rubric shifts, and corrective actions. 2. Client Partnership & Governance Act as the primary interface with frontier stakeholders. Run structured sprint reviews and executive updates. Surface risks, trade-offs, and improvement hypotheses proactively. 3. Quality & Rubric Governance Drive guideline version control and calibration systems. Implement layered QC (SME review, audits, gold sets). Lead root-cause analysis for model performance issues. 4. Team Leadership & Capability Building Lead TLs, SMEs, QA, and large annotation teams. Ensure structured onboarding and continuous calibration. Monitor cognitive load and prevent quality fatigue. 5. Metrics & Profitability Control Track throughput, rework, guideline drift, and SLA adherence. Maintain cost discipline across iteration cycles. Protect margin during rapid scale transitions. 6. Risk & Compliance Oversight Enforce data confidentiality and IP protection. Mitigate bias risk, contamination risk, and safety mislabeling. Required Competencies Strong governance mindset with structured thinking in ambiguous situations. Working knowledge of SFT, RLHF, evaluation pipelines, and safety labeling. Ability to convert evolving guidelines into scalable workflows. Executive-level communication and structured reporting capability. High ownership, accountability, and ethical judgment.