Replatforming a global industrial PC maker for composable commerce.

The legacy monolith turned every business change into a workaround, and three problems compounded it.

First, a complex catalog. OnLogic's products are configurable to thousands of variants: different chassis, processors, memory, and I/O combinations. The legacy system flattened this, forcing customers through a buying experience that felt more like a database query than a configurator.

Second, a global audience. Multiple languages, currencies, tax regimes, and localized content, none of which the existing platform handled gracefully.

Third, a constrained roadmap. Strategic projects kept ending with "we can't do that on our current stack," and innovation cycles stretched out.

We rebuilt the commerce stack around three principles: composable, headless, swappable.

commercetools became the headless core, handling cart, checkout, and order orchestration. Contentful took on content. Algolia owned product discovery and search. Each piece was selected for what it does best, and each can be replaced when something better comes along.

I ran a co-delivery model across both teams: Orium engineers and OnLogic engineers, spanning North and South American time zones, working as one squad. We started with a proof-of-concept to validate the architecture against OnLogic's most demanding flows before scaling commitment. As delivery manager, I owned the plan, the budget, and the client relationship, keeping OnLogic's leadership close to every decision.

OnLogic now operates on a platform built for their business. A custom price engine handles configurable products without manual reconciliation, and checkout supports the payment methods and tax rules of every international market it serves.

The team also came out of the project with a platform its roadmap can build on, so new initiatives no longer stall on platform limits.