Build a centralized experiment operations tool that supports A/B testing, multivariate tests, Bayesian analysis, causal inference, DOE setup, offline model evaluation, and post-experiment reporting. The product would provide templates for experiment design, integrations with analytics and data warehouses, evaluation-set tracking, automated significance and impact analysis, and shared reports for roadmap and strategy decisions.

Job signals show experimentation becoming a core operating layer across personalization, discovery, AI evaluation, GTM^?, and manufacturing optimization. The rise of LLM^? and personalization systems increases the need for rigorous offline and online evaluation frameworks tied directly to business impact.