AI for Maximo: seven questions
to ask any supplier

More and more suppliers now offer an AI assistant for IBM Maximo. They all promise the same thing: ask Maximo a question in plain English and get an answer. The differences only show up once you look underneath.

Before you choose, put these seven questions to every supplier on your list, including us. Each one comes with a line on why it matters. Our own answers are underneath each question, so you can hold us to them too.

1. Which Maximo versions does it run on?

Why it matters: Many sites run Maximo Application Suite 9, many still run Maximo 7.6, and some run both. An assistant that only works on the newest release may mean an upgrade project before you see any value.

Our answer: Interpretos runs on Maximo Application Suite 9 and on Maximo 7.6, through Maximo’s standard interface. No upgrade, no plug-in and no change to your database is needed.

2. Does it give the same answer when asked twice?

Why it matters: AI models can phrase things differently each time. That is fine for a summary. It is not fine for a backlog figure that goes into a management report.

Our answer: Every figure comes from a query run against your live Maximo data, and the working is shown with the answer. If two answers ever differ, you can see exactly why. Every change of AI model is gated on a test battery before it reaches you.

3. Can you see the records behind every figure?

Why it matters: A number nobody can check is a number nobody should act on. Your planners will want to click through to the work orders behind it.

Our answer: Yes. Every answer shows how it was reached, so a user can check it in Maximo. Every question, and every query made against Maximo, is also logged on your own system for your auditors.

4. Does it say so when it does not know?

Why it matters: The most dangerous answer is a confident wrong one. You want an assistant that tells you when it is unsure, not one that fills the gap.

Our answer: Every answer shows how confident Interpretos is. When an answer stops short, the user can press “Dig deeper” to have it go further.

5. Does it respect each user’s Maximo permissions?

Why it matters: An assistant that reads Maximo through one all-powerful account can show someone costs, or another site’s records, that Maximo itself would hide from them.

Our answer: Yes. Each user connects with their own Maximo credentials, kept separate and encrypted, so each person sees only the data their own Maximo access allows. Interpretos only reads. It never creates, changes or deletes a record.

6. How does it learn your own rules and terms?

Why it matters: No two Maximo sites are set up alike. Your own statuses, fields, site structure and shorthand are where generic AI goes wrong.

Our answer: Interpretos arrives knowing Maximo and learns your site from the people using it, with human sign-off on everything it learns. In a ten-week trial at Stena Drilling it learned 26 Stena-specific rules this way. Read the case study.

7. Where does your data go?

Why it matters: Your maintenance data, costs and supplier prices are commercially sensitive. You should know exactly what leaves your network, and to whom.

Our answer: Interpretos runs on your own infrastructure. You choose where the AI runs: a provider of your choice, your own cloud tenancy, or a model on your own hardware with nothing leaving your network at all. Your Maximo credentials are never sent to the AI. See exactly what goes where.

Put the questions to us first

We would rather you test us than take our word for it. Try Interpretos on a copy of your own Maximo data and ask it the questions your team asks every week.

IBM and Maximo are trademarks of IBM Corporation. Interpretos is made by Code Development Ltd, the company behind MaxTAF, and is not affiliated with IBM.