Technical articles, guides, comparisons and field reports on what companies actually do with AI once it is deployed. We write from the engagements we run: more than 40 AI diagnostics in under a year, across estates where Claude, ChatGPT, Gemini, Copilot and in-house agents coexist.
Measuring the real usage of AI platforms from the logs a company already produces. Arbitrating an AI portfolio: what to industrialise, what to consolidate, what to cut. The European AI regulation and what its deadlines change in practice. Agent runtimes and the open frameworks we build on, LangGraph, Google ADK and MCP. And the recurring finding of our work: across the estates we analyse, 20 to 40 % of the AI licences paid for go unused.
Every article carries a named author from the Idun Group team, the studies it relies on, and its publication and update dates. We cite sources rather than assert. No pricing is published on this site: every quote follows the analysis and is given in a meeting.
