 AI has such a big glamour that everybody follows into it. We should have it. I mean, there is something I don't have, I should. And I think a business never, this is kind of obvious, but never forget that the main thing is his business, not the technology. So before jumping into complex algorithms, I think every business has to make a homework. Point out the core business, the main KPIs, and use very simple methods, very simple algorithms. Maybe sometimes in order to do forecasting, you don't need AI, you just need an organized system, digitize the entire supply chain, for example, and just moving average, for example, or polynomial regression, that's enough. And that would give the companies a very good starting point and create a benchmark against which they'll be able to compare AI in the future, for example. And it will give them a safe ground upon which they can stand and compare future work in advance. That's business. Absolutely business. And it's not related at all with the algorithms. For example, if you have a forecasting system, very simple. You need to see how well it did, but how well it did concerning business. For example, if you have a company that's handling 10,000 SKUs, you have a forecasting that's going to forecast 10,000 SKUs. But your business, it could be very well possible that your business is managed or controlled by 200 SKUs. And 80% of all your revenue comes from it. So what you're really concerned about is about forecasting those 200 SKUs. So you have to keep business in mind all the time and not get mislead by the KPIs of algorithms, machine learning, whatever. You have to be very careful about that because technical people, we tend to go for the technical measurements which are not relevant at all with the business.