The big shift is stability and performance. Previously, some of our core services were running in an environment that wasn’t optimal – Windows-based nodes that limited us. We’re now upgrading everything to the latest .NET and moving the entire Norce Commerce engine into Kubernetes clusters running in Linux.
That means our partners will experience the platform exactly as intended: much more stable, and as a bonus, about 40% faster in our internal tests. The goal is a silent transition – the same endpoints and functions you’re used to, but with far better reliability and performance behind the scenes.
Norce Admin is already running on the new platform. Most partners won’t have noticed because we’ve rolled it out gradually and invisibly. The big piece – what we internally call the Commerce Service, is now in final testing. This is the core API that our partners build their e-commerce solutions with and one of the most essential parts of our platform. It needs to be fast, stable, and always available – because it directly impacts what customers see and do in their shopping experience. We expect to push it to production within the next few weeks.
Then we have Connect Service which is the integration API used to push data into Norce Commerce. It’s the engine behind syncing product information, prices, stock levels, and other business-critical data from external systems into the Norce ecosystem. After that, we’ll continue migrating our remaining services, one by one, until the entire stack runs on the modernised platform.
For our engineering team, the difference is huge. In the old environment, we were limited and forced to build defensively. Now, with the latest .NET and Linux/Kubernetes, we can develop much more offensively – faster iterations, faster improvements.
For our customers and partners, the experience will feel familiar – intentionally. The interface and functions stay the same, but under the hood it’s a completely different engine. The result is greater stability and performance without forcing partners to make changes to their own systems.
In the short term, partners can keep working as before – their storefronts and integrations won’t break. That was a key requirement.
Once all our core APIs are fully modernised and running the way we want, our top priority is to build new features. Modernisation has taken a lot of time, and our partners have been patient – now we’ll be able to deliver much faster.
AI has become a truly valuable part of how we work. One of the biggest advantages of modernising our stack is that it enables us to use modern AI tools for both development and testing – which helps us deliver faster, with better quality.
Lately, we’ve started working closely with Devin.AI, a code generation tool that helps us with everything from fixing bugs to handling complex migration tasks. It works continuously in the background – suggesting changes in small pieces – and then we use GitHub Copilot to review and fine-tune the code.
In practice, this means less repetitive work for developers, more time for innovation, and an extra safety net in our delivery.
Devin and Copilot have basically become part of the team. And yes – we’re still waiting to see if they show up for after-work drinks!