Experimentation Systems

Causality at Scale

True product-led growth requires a culture of rigorous, high-velocity experimentation. But as companies grow, bad tests, peeking at p-values, and statistical illusions inevitably corrupt decision-making.

During my tenure at Zapier, I established the foundation of an experimentation culture that empowered engineers and product managers to safely ship changes directly to millions of users. I bring that exact enterprise-grade rigor to growing startups looking to scale their own A/B testing velocity without sacrificing statistical integrity.

Experimentation Capabilities

• ✓
  Platform Integration: Setting up modern feature-flagging and assignment engines (e.g., LaunchDarkly, Statsig, GrowthBook) across your entire stack.
• ✓
  Statistical Frameworks: Implementing fixed-horizon testing, sequential analysis, and proper alpha-spending to allow safe, continuous monitoring.
• ✓
  Variance Reduction: Utilizing CUPED to dramatically shrink confidence intervals and slash required sample sizes.
• ✓
  Cultural Transformation: Coaching product teams on power analysis, sample sizes, and how to define primary, secondary, and guardrail metrics.