A closer look at some projects I've worked on. The challenges, approach, and results.
Regional construction contractor — A 200+ person contractor running multiple active worksites.
Day-to-day operations relied on spreadsheets and group chats. As projects got bigger, small coordination issues turned into expensive mistakes and delayed billing.
Read case studyCombo — French HR SaaS for restaurants and hospitality (~130 employees, 10,000+ customers)
Combo's sales team juggled five tools for every prospect: Google Maps, Instagram, business registries, Salesforce, Looker. Each lead meant starting from scratch. Want to find "restaurants in Lyon with strong Instagram and multiple locations"? That required opening five tabs and cross-referencing manually.
Read case studyCombo — French HR SaaS for restaurants and hospitality (~130 employees, 10,000+ customers)
Combo had 124,000 restaurant listings from Google Maps and 277,000 from French business registries. No shared identifier between them. A Google listing for "Chez Marcel" might match "SAS Groupe Restauration Marcel et Fils" in the registry—or it might not. Without linking the two, neither dataset was useful for prospecting.
Read case studyCoCo.ai — WhatsApp marketing and cart recovery tool for Shopify stores, built on AI. Small team, about 10 people. A venture firm bought them to grow from a few hundred merchants to several thousand.
A venture buyer was acquiring CoCo.ai and needed to understand where the product stood technically before closing. CoCo had been built fast by a small team, which is normal for a startup at that stage. The buyer's question wasn't whether the code was perfect. It was: what's the real foundation, what will scaling cost, and what should the team prioritize after the acquisition?
Read case study