Data Readiness, Learn by Building, and the Build-or-Buy Decision

I think sometimes we lose sight of that when we buy technology and we build technology solutions.

There's kind of a feeling that like, oh, you know, like it's working and it'll keep working and I don't need to touch it.

I feel like that opening line for every one of them is this

changes

everything.

So I'll let you talk on this one a little bit because I've, I've peppered it throughout, but so how do all those three things work? relate to a mid-market company who's just starting on this journey? How does...

The one practical thing that companies should start doing to kind of tackle these issues? I think a big piece of

it, I mean, there's two, right? I think that when it comes down, it's the data, right? But not database data, right? Right? Like this is this giant pile of everything stored in the cloud and I never throw anything away.

Right.

I think that's what people don't understand.

And I'm stopping here again because the data is what you always hear.

We've got to get our data ready for AI.

We're not ready for AI because we don't have our data ready.

They're not talking about your operational data warehouse that you've been using for 20 years.

They're talking about all the other stuff that adds the color to that financial or operational data.

I just want to be really clear on that distinction because that's what people mean.

How do we get all the SharePoint version 12s, you know, sorted through?

And I think, one, having that either cleaned up, curated, or marked, and then building systems, not just telling the LLM, go read all this stuff, right, and try to help me out, but getting smart about the systems that we built.

We tagged our data pretty specifically, and now we're learning that we need to do it a little bit more specifically, but, right, conversations are... an object but a human conversation and an llm conversation both valuable but you know need to be um marked as such so just in layman's terms like because you use the word object

and yeah but like you've no no it's okay i just i want to make sure everybody kind of understands what we're talking about so You've got to really decide how to categorize all of this data.

And I think that is probably one of the biggest planning pieces that are hanging people up right now, which is, well, how do I model all this out and say, well, this is this and this, I got to put this in this bucket and this in this bucket and this in this bucket.

And those buckets all have to be connected with a pipe.

And so that's the real issue that we're trying to solve.

And I think what people mean about the data, but you also had a second point.

So I think we got stalled out there.