You Don't Need to Prepare Your Data for AI

Alex GruebeleJuly 21, 20266 min read

"We know AI can probably be used to solve some of our problems. But our data isn't ready for AI. We need to get our house in order, then we can launch AI projects."

I hear something like that from a manufacturing exec a few times a week. Careful is the right instinct with new technology. But this particular worry usually rests on a picture of AI that's a few years out of date. When we dig into the problems the client wants to solve, I'd say 9 out of 10 times the data is ready as it sits.

The reality is that modern AI (and in this article, I'll refer to AI as the latest form of artificial intelligence models, large language models) is well suited for making your data "AI ready". It's incredibly efficient at turning unstructured data into structured data for the vast majority of tasks a human would do by hand. Getting your data "AI ready" is itself a job that AI does well. You don't need a two year data cleanup program before the first project. The first phase of most projects is the cleanup. The technology you're waiting to be ready for is the same technology that gets you ready.

Industrial Executive Briefing

Reach out for access to all Tau9 Executive Briefings