Quality Analytics Lead

Quality Data Modeling & Analytics Infrastructure Design, build, and maintain dbt models and data marts that serve the Quality organization’s enterprise reporting needs — covering throughput, accuracy, defect rates, CAPA effectiveness, annotator/rater performance, and program-level quality health. Use Python for higher-order data modeling tasks including cohort analysis, performance trend modeling, and custom aggregations that go beyond standard SQL/dbt scope. Partner with data engineers to define source data requirements, document data lineage, and ensure quality data is reliable, consistent, and analytics-ready.

Own the quality analytics data layer end-to-end: from raw operational inputs to clean, tested, well-documented marts consumed by dashboards, reports, and ad hoc analyses. Apply dbt testing, documentation, and best practices to build a trusted, maintainable codebase that scales as new programs and data sources are onboarded.

Quality Measurement Frameworks & Metrics Design. Collaborate with Quality Managers and Analysts to define, standardize, and operationalize quality metrics — including accuracy rates, defect categorization, sampling coverage, inter-rater agreement, and CAPA closure effectiveness — consistently across all programs. Design measurement frameworks aligned to acceptance criteria and quality thresholds, ensuring metrics faithfully reflect program health and client commitments. Support rubric and guideline effectiveness measurement, helping quality teams understand whether their standards produce consistent, measurable outcomes across annotators and raters.

Champion data quality governance within the Quality org: own metric definitions, threshold documentation, and analytical methodology standards to reduce inconsistency and reporting variance. Define enterprise-level quality dashboards in partnership with BI resources, translating mart output into clear, decision-ready views for Quality Managers through to senior leadership. Analyze patterns in model evaluation outcomes, annotator disagreement, and guideline interpretation to surface systemic issues in AI training data and evaluation processes.

Experimental Design & Performance Validation. Design and execute A/B tests and controlled experiments to measure the impact of quality interventions, process changes, and annotator training programs — applying proper power analysis, significance testing, and results interpretation. Build success validation frameworks to confirm that CAPA actions and process improvements produce measurable, sustained outcomes — not just short-term fluctuations. Develop performance attribution models that quantify the contribution of specific quality initiatives to outcome improvements, separating causal signal from noise in program performance trends.

Apply statistical methods to sampling design, audit analysis, and error pattern detection, surfacing systemic quality issues and their root causes with data-backed evidence. Conduct pre/post analyses for major quality program changes, training rollouts, and rubric updates, delivering clear impact assessments to quality leadership and clients.

Decision Support & Stakeholder Partnership. Act as the analytical partner to Quality Managers (P2–L2) and senior quality leadership, translating complex data models and analytical findings into clear, actionable insights for program decisions. Produce client-ready analytical deliverables — including quality performance summaries, trend analyses, and post‑mortem reports — that Quality Managers can present in client governance reviews and executive forums. Proactively monitor quality performance data to identify emerging risks and flag issues to quality leadership before they escape into client-impacting problems.

Lead discovery conversations with quality stakeholders to understand their data needs, translate them into well-scoped analytical requirements, and ensure delivered solutions are aligned with actual decisions. Coach quality team members on data-driven decision making — helping them frame analytical questions, interpret results, and embed measurement into their processes from the…