An e-commerce that runs itself
Built a coffee e-commerce from zero, then automated its operations end to end — restocking, support, pricing and anomaly detection.
Problem
A coffee e-commerce had to be built from scratch and made into a real, profitable business — then kept running without a large team behind it.
Approach
I owned everything: Shopify CMS/CRM, Google Ads, brand identity and customer service. Then I built the data engine behind it — price-elasticity and demand-forecasting models — and an AI platform automating restocking, customer support, dynamic pricing and supply-cost anomaly detection.
Result
CHF 500K+ first-year revenue, days-out-of-stock cut to zero, and +70% campaign engagement via A/B testing. After returning to rebuild degraded support and email marketing, a +30% increase in orders.
Notes
Effectively the AI/data engine of a live P&L — the most differentiated thing on this site, because most data scientists never ship automation into a real business they’re accountable for.
Screenshots coming soon.