All speakers

Joseph Powers

Principal Data Scientist, Intuit

Joseph is a Principal Data Scientist at Intuit, where he has lead Experimentation and Advanced Analytics. He is highly interested in experimentation, designing for large effects, simulation studies, Bayesian modeling, optimal stopping and decision theory. He will be very glad if you approach him to discuss any of these topics during the conference. Previous careers include 7 years as a custom furniture maker and a Psychology PhD from Stanford. When he is not programming or playing dad he enjoys woodworking, cooking and surfing.

Joseph Powers

Sessions / 2025 / 2 talks

  • Going Bayes: Shifting our Testing Methods to Reflect our Priorities

    Joseph Powers shares insights from implementing Bayesian Risk-Based Testing at Intuit, exploring how prioritizing business outcomes over error rates reduced test duration by over 60%. This talk offers valuable perspectives on making decisions based on potential losses rather than significance testing and scaling acceptance across analytics, product, and leadership teams, particularly valuable for practitioners running A/B tests or driving organizational change in experimentation.

  • Going Bayes: Shifting Our Testing Methods To Reflect Our Priorities

    Bayesian AB Testing at Scale: How Intuit Revolutionized Experiment Design | Discover how Intuit transformed their experimentation framework using Bayesian risk-based testing to achieve 60% faster results. Learn practical implementation of risk threshold algorithms that optimize for business outcomes rather than traditional error rates. Master strategies for organizational adoption of advanced statistical methods across Analytics, Product, and Marketing teams. Features detailed case study of successful enterprise-wide statistical transformation, including implementation challenges and measurable outcomes.

Ready to take this stage?

The next edition is programmed by practitioners. Tell us what you built and what you learned.

Apply to be a speaker