of Italian companies with at least 10 employees use AI in 2025, up from 8.2% in 2024
Ask the AI to analyze all your maintenance data
Natural language questions. As easy as asking a colleague.
MACHINA AI Link lets artificial intelligence analyze MACHINA WEB data: which maintenance tasks are overdue or how a similar failure was fixed. The AI answers by analyzing your data.
Contact Us How it worksAI in business is already a reality
In just one year, the number of Italian companies using artificial intelligence has doubled. Those who use it to analyse their own data make better-informed decisions; those who wait risk falling behind.
of large Italian companies already use at least one artificial intelligence technology
of respondents to an international survey say AI helps them make better decisions
Among companies that considered AI without adopting it, 45.2% cite the availability or quality of data as an obstacle. With MACHINA WEB, maintenance data is already collected in a structured way and linked to machines, interventions and spare parts: a solid starting point for AI.
Sources: Istat, Enterprises and ICT – Year 2025; McKinsey, The state of AI in 2026.
Maintenance data is a company asset: with AI, consulting it is even simpler
Year after year, MACHINA WEB builds up a wealth of information: interventions, failure causes, technicians' notes, working hours, costs and spare parts used, already organised into reports, analyses and KPIs.
Artificial intelligence adds an even more immediate way to consult it: you ask the AI a question as you would ask a colleague, and the AI searches, analyzes and offers an answer.
To get started, simple questions like these
- "Which maintenance tasks are overdue?"
- "Which failures are keeping a machine down?"
- "Which spare parts are below minimum stock?"
- "How much did contractor labour cost last quarter, broken down by contractor?"
- "Which open intervention requests concern critical assets?"
In the main warehouse, 4 spare parts are below minimum stock:
| Item | Stock | Minimum stock | Used on |
|---|---|---|---|
| Hydraulic oil filter | 0 pcs | 2 pcs | Presses P-02, P-04 |
| Bearing 6205-2RS | 2 pcs | 6 pcs | Pumps PMP-01, PMP-03 |
| SPZ 1250 belt | 1 pc | 4 pcs | Fan VNT-02 |
| 10 A fuse | 3 pcs | 5 pcs | Robot RBT-01 |
Warning: the hydraulic oil filter is out of stock and is required by 2 work orders next week. It should be reordered first.
Sample conversation: data and answer format are for illustration only.
The solution to a failure is often already written in the notes of a past intervention: with AI you find it in no time, instead of starting the diagnosis from scratch.
Claude, Copilot, ChatGPT or an in-house AI
MACHINA AI Link follows MCP (Model Context Protocol), the open standard AIs use to connect to company data, and does not tie you to any AI vendor: you can use the service your company has already adopted, or switch in the future without changing anything in MACHINA WEB.
Online AI
Claude by Anthropic, Microsoft 365 Copilot, ChatGPT by OpenAI, Gemini Enterprise by Google and other services compatible with MCP servers connect to MACHINA AI Link over the Internet, via HTTPS. They are the recommended choice for complex questions and analyses on large amounts of data.
In-house AI (local AI)
With an AI installed on a company computer or server, for example with LM Studio or with Ollama and Open WebUI, questions and data stay on the company network. It is the choice for those who do not want data to leave the company or to depend on external services.
With AI Link, Studio ARCHIMEDE makes AI easier to use
Using artificial intelligence to analyze maintenance data is a project in its own right.
MACHINA AI Link is built by Studio ARCHIMEDE to give the AI the meaning of all the data managed.
AI Link can be installed by your company's IT department or, on request, by Studio ARCHIMEDE.
- Designed for maintenance
- Every piece of data explained to the AI
- Step-by-step guides to connect each AI
- Continuous improvement
No reports to build, no filters to set: write a question and the AI finds the answer in your maintenance data.
What an MCP server is and why the AI needs it
On its own, AI does not know your plants, your technicians or your interventions. MCP (Model Context Protocol) is the open standard AI uses to connect to a program and read its data. MACHINA AI Link is the MCP server of MACHINA WEB: it bridges the AI and the maintenance database.
Search tools
The tools are over 30 predefined searches that the AI can run on MACHINA WEB to find assets, intervention requests, completed interventions, maintenance plans, work orders, technicians, items, warehouses and purchases. The AI decides on its own which ones to use and with which filters, even several in a row for more complex requests.
The AI learns maintenance
Data alone is not enough: a general-purpose AI does not know the codes, statuses and rules of industrial maintenance. MACHINA AI Link does not just hand it a table of data: it teaches it the meaning of each piece of information and how to use it, so the AI reasons according to maintenance logic instead of guessing.
Ready-made prompts
Guided questions, designed by Studio ARCHIMEDE and tested on MACHINA WEB data: just enter one detail, such as a request number, and the AI follows a predefined path on which data to read and how to present it. The result is a more complete answer, always in the same format.
From your question to the AI's answer in four steps
The AI never accesses the database directly: it asks MACHINA AI Link for the data, which reads only the permitted information and returns it together with an explanation of its meaning.
Your question
The user writes to the AI in their own words, giving what they know: an asset, a department, a time period.
Choosing the tools
The AI decides which MACHINA AI Link tools to use, with which filters and in which order.
Reading the data
MACHINA AI Link reads the data in MACHINA WEB and sends it to the AI explaining its meaning.
The answer
The AI cross-references the data received and writes the answer: a list, a table, a summary or a recommendation.
Example: "How can I fix the failure in request 123/26?"
A single question can lead the AI to use several tools in a row. With the MACHINA AI Link ready-made prompt, the path is:
- the AI reads the intervention request, its notes and the asset data;
- it looks for similar failures already fixed on the same asset or on similar assets and reads the technicians' notes, which often say what they found and how they fixed it;
- it lists the technicians who worked on the similar failures, who can be asked for advice;
- only afterwards, if it can, it searches the Internet for suggestions, citing its sources.
The AI consults the data but cannot change it: access can be limited to a single company or a single plant, and every request it makes is tracked in the activity log.
You decide who sees what
Data access is governed by keys created in the MACHINA AI Link administration area: one for each AI or group of people, each with its own limits.
One key per group
You create a key for each department or office and choose which data it can read: all companies, a single one or a single plant. Each key can have an expiry date and, if needed, can be deleted in an instant without disconnecting the others.
Activity log
Every AI request is logged with the tool used, number of results, duration, outcome, key and originating AI program. The log is encrypted and can be consulted from the administration area.
Protected access
HTTPS connection with OAuth 2.1 authorisation or an access key, an IP address filter and two-step verification for the administrator: only authorised programs and people can reach MACHINA WEB data.
Questions about MACHINA AI Link
Works with the AI your company already uses, or with a local AI.
Bring artificial intelligence into your maintenance
Contact us to find out more about MACHINA AI Link.
We will explain how the AI can answer your department's questions starting from MACHINA WEB data.