Your personal assistant
Ask in plain Arabic or English, the way you would ask a colleague.
Every user gets a personal assistant that answers in plain Arabic or English from your own data. It only sees what that user is allowed to see, and every answer comes with its sources.
For a large organisation, the question is not whether AI can answer, but whether the answer is right and allowed. The assistant is built for both.
Ask in plain Arabic or English, the way you would ask a colleague.
Questions go to the right specialist: assets, inspections, sensors or reporting.
The assistant works with each user’s own permissions. It cannot see what they cannot see.
Every answer is checked against the data before it is shown, with its sources.
Connect your own Claude or ChatGPT to your data securely, with the same permissions.
Set how much each role can use, and keep keys and settings secure.
Follow a question from an asset manager. Click a step, or let it play.
"Which bridges in the Eastern Region are overdue for inspection, and do any have urgent defects?"
The question is recognised as an inspection question and handled by the inspections specialist.
The assistant looks up overdue inspections and urgent defects, as the user, in her region.
The numbers in the answer are checked against the data: 7 overdue, 2 with urgent defects.
"7 bridges are overdue, 2 with urgent defects", with a table and a link to each inspection.
Example with sample data.
Common questions are recognised straight away; the AI is used only when needed.
Each specialist handles one area, so it answers with focus.
Figures come from your data, under the user’s own permissions.
The answer is checked against the facts before it is shown.
There is no all-seeing account behind the assistant. A regional engineer gets regional answers, an executive gets the whole picture, and a contractor gets only what a contractor may see.
Accuracy comes from the design: focused specialists, figures calculated from your data, and a check before every answer. It does not depend on the most expensive model.
Connect Claude or ChatGPT to your asset data through a secure sign-in. They see only what the user may see, and they can only read.
The value of AI is getting to the right answer quickly. Mintema gives you the speed without losing control of your data.
Questions that needed an analyst and a report are answered on the spot.
People who never open a dashboard can still ask, in their own language.
Permissions, read-only access and full logs apply to every question.
Use the built-in assistant or connect the AI your organisation prefers.
Have a question that is specific to your organisation? A solution engineer will answer it against your own assets, processes and data.
No. It reads and explains. It cannot change anything.
Yes. It works with each user’s own permissions and organisation.
Yes. They can connect securely to your data, with the user’s permissions and read-only access.
Yes. Ask and get answers in Arabic or English.
Bring the five questions your managers ask every week. We will put them to the assistant on sample data during the call.