Is our data ready for AI?

Eighteen statements in plain language, about five minutes. Score your organisation honestly. You will see whether your data can carry an AI application, where it stalls, what that costs and which steps come first.

First: what matters most to your organisation right now?

Your result will be framed around this goal.

Steer

Does management know what it wants AI for, and what that asks of the data?

We know for which concrete questions or tasks we want to use AI.
For every AI application we know which data it needs.
Management understands that the outcome of AI is as good as the data behind it.
Someone in management is accountable for the data we use in AI.
We have agreed which data we do and do not put into AI applications.
We know what it costs us when an AI application gives a wrong answer.

Build

Is the data set up to be used by AI?

For the data AI uses, it is clear who decides what is right.
The meaning of that data is recorded, so that people and model mean the same thing.
We know how good the quality of that data is, and whether that is enough for this purpose.
Customers, products and other shared data exist once, not five times.
We know where that data comes from and what was done to it on the way.
It is settled who and what has access to which data, for AI applications too.

Run

Does it stay reliable once it is running?

For an AI answer, we can trace which data it is based on.
People check AI outcomes before a decision is taken on them.
When the data changes, we notice before the AI application gives wrong answers.
Errors in AI outcomes are reported and traced back to the data.
Employees know which data they may and may not enter into AI tools.
We can show what AI has delivered for us, and what the data contributed to that.
0 / 18

I’d love to hear from you