In the second quarter of 2026, Italian machine tool orders fell 25.8%, yet production and revenue hold up thanks to the order book. A Vedrai Observatory report on how to read demand in advance and prepare operational choices in capital goods, before the change reaches the books.

On November 4th, Vedrai takes the stage at the Final Conference of the Data & Decision Intelligence Observatory hosted by Politecnico di Milano's.

Price lists update once or twice a year while costs move constantly: every offer risks starting with margin already eroded, and sales has no way to notice. It is not a people problem, it is an information problem and it is solved at the moment of negotiation.

A forecast gives you one number. Business scenario simulation shows the full range of outcomes for every option you could choose, with its effect on margin and its risk, before you commit capital or time.

Making better decisions with AI isn’t about getting an answer faster; it’s about making every decision modelable, comparable and measurable before you act.

When competitive advantage is funded by industrial margin, how you buy the raw material is not a purchasing detail: it is the lever that defends the whole business model. Here is how to decide price, timing and FX without tying up too much cash.

For a mid-sized firm a single investment can be worth a third of its capital, yet it is decided on one number. Three questions to commit it on a probability, not an estimate.

In fashion and sport retail, goods are bought months ahead of uncertain demand: buy too much and you end in sales, too little and you lose sales. The decision that matters is not volume, but where to concentrate shelf exposure, with a risk map.

Acquisitions, new products, new markets, divestitures: in 2026 the assumptions behind corporate plans held for less than four months. How simulation, continuous monitoring, and AI turn the plan from an estimate to defend into a decision to govern.

Retail and manufacturing: from the 2027 headcount plan to week-by-week scheduling. Why budget and shifts must be decided together.

AI and agentic AI do not create the knowledge a company decides with: they use it. The real prerequisite for applying them to critical decisions is not a more powerful model, but three layers to build first: formalized decision knowledge, a semantic layer and a decision layer.

When the result closes below budget, the easy explanation is the market. Often the damage is concentrated in very few jobs. How to see where value was lost, how exposed the pipeline is to steel prices, and how to recover while holding cash.