Ansel·May 10How I Measured the Real Impact of a Global Product Launch Using Synthetic ControlA practical, scalable causal inference workflow for real-world e-commerceA response icon1A response icon1
Ansel·Apr 14When High-Propensity Users Are Not the Best EDM Target: Building an Uplift Modeling Framework for…Note: All numbers, charts, and examples in this article are illustrative and based on demo-style outputs that preserve the analytical…
Ansel·Mar 17Designing Cross-Sell Recommenders for High-Propensity UsersDebiasing popularity and improving category diversity in EDM recommendations
Ansel·Mar 11A/B Tests Don’t Make Decisions. People Do.From Statistical Significance to Structured Decision Logic
Ansel·Jan 14Do High-Propensity, Recently Active Users Actually Need EDM to Convert?All numbers in this article are illustrative, derived from a demo dataset that mimics the real patterns but does not expose any actual…
Ansel·Dec 10, 2025From Generative BI to a BI AI Agent: Orchestrating Multiple Data Sources with AWS AgentCoreHow I used AWS AgentCore to let multiple data sources and PDFs talk back to the business
Ansel·Nov 27, 2025How I Built a Generative BI Prototype That Lets PMs and Marketers Talk Directly to DataHow I used AWS’s Generative BI framework to let PMs and marketers query e-commerce data without SQL
Ansel·Nov 12, 2025Evaluating Cross-Cloud Personalization Systems: A Practical Experiment Comparing AWS Personalize…Introduction: Why Personalization Matters More Than Models
Ansel·Oct 24, 2025From RFM to GMM: How I Built a Customer Segmentation Model That Product and Marketing Teams Can…Not just building another model — but delivering a strategic asset that product teams can actually use. In this post, I’ll walk you…
Ansel·Oct 13, 2025Phase 3 — From Self-Correcting Models to Measurable Impact: Designing A/B Tests for ML SystemsPart of my end-to-end ML system series — proving real business impact through A/B testing