We design, build, and run commerce platforms, custom applications, and integrations for online businesses across the Nordics. Senior people stay close to the work, from first decision to long-term support.
We bring AI to commerce work when it earns its keep. Chatbots that recommend, classification that scales, integrations that turn public data into commercial leads.
AI is one tool among several. We use it where it shortens a real workflow or solves something the platform couldn't do otherwise. We don't add AI features so we can say we use AI.
We start by understanding the problem. Sometimes the answer is AI; sometimes the answer is better data, a cleaner workflow, or a smaller change to the platform. The right answer earns the engagement.
We use OpenAI, Anthropic, and other providers behind our own integration layer. Switchable models, observability, retry and fallback paths. Not vendor-locked to a model that might change tomorrow.
AI we ship lives inside commerce platforms, custom applications, and operational workflows. Product recommendations, customer support, data enrichment, lead identification, catalog classification.
AI applied to a real catalog problem: classifying long product codes into structured categories. Reduced a manual taxonomy effort from weeks to hours, with rules and review built into the workflow.
A chatbot and interface connecting to the Norwegian Matrikkelen (the national cadastral API). Helps developers identify land lots with growth potential. Public registry data turned into commercial workflow.
No. We are a commerce-platform partner that ships AI features when they earn their keep. AI is one capability among several, not the pitch.
OpenAI, Anthropic, and others depending on the workload. Embedded behind our own integration layer so models are switchable and operations observable. We pick the model that fits the task, not the one with the loudest marketing.
Yes, when we can describe what the AI is supposed to do in concrete terms. Custom chatbots, classification, content workflows, lead identification. We won't take "we want AI" as a brief; we will take "we want to reduce manual catalog work by half".
Per-client data lives where it always lived: your stack. We do not send catalog or customer data to model providers without explicit permission, and we set up data boundaries when we do.
Yes, regardless of platform. We have integrated AI features with Magento, custom Next.js storefronts, and B2B portals. The platform changes; the integration discipline stays the same.
