Mm.

From one end, the industry has a lot of opportunity, but there's also a lot of change, and competition is picking up.

On the, on the other hand, there's new technology that can actually optimize and increase productivity.

So, you know, w- we're, uh, in an environment where brands need to really come up with the right product to get placed on the shelf.

Uh, retailers have so many home brands that, you know, may provide the same experience at a lower cost, so that pushes brands to, uh, a point where they really need to get closer to the consumer.

They need to ship products faster and, more importantly, they need to ship them very accurately.

The same time, we're seeing mega shifts like GLP-1 impacting food consumption, especially with the higher income consumers that have the, you know, availability to, uh, get, uh, on GLP-1 and, and use the drug.

Um, so I, I want us to think a little bit about top-down versus bottom-up change, and when those two forces meet, what's the best thing we can learn, uh, from, from that merger?

Why is this moment different from five years ago or even different from two years ago? Because everything's changing so fast.

Well, I think that the answer is that everything is changing so fast.

This is actually the short answer.

Now, the long answer is that I started building AI for food and beverage back in 2017.

2018, we already established a company and we've been working with thousands of brands.

But what we're seeing today is that the pace of change is so robust And it's happening from outside the food industry, right? It's happening where LLMs and the foundational tech actually started with consumers and only then moved to enterprises.

So expectations are different all around, right? So consumers have different expectations because when they're scrolling today on their feeds, everything is so much more hyper-personalized to their needs, sometimes overly personalized to their needs, right? And, you know, some interesting facts is that consumers believe more agentic influencers than real influencers when they come to do like a real purchase decisions, because at least they're not feeling that they're being, you know, manipulated because they saw this is actually someone who is an agent.

What are some keys to taking that to a pilot and then to like making that real decision? What are the questions that teams should be asking themselves about the data maybe or about the insights and how to use those?