If not, we’ll be happy to do it in the upcoming or the next analyst visit that we have at the next Investor Day is to show you what happen when you run certain scenarios through Enlighten versus any kind of generic AI solution, and the difference is night and day. And our customers, many of them, already try to deploy generic AI in their environment. And this is exactly the point that they encounter, the issue of precision but precision in context, and that’s the reason why they are selecting Enlighten. So this is what we’re doing. Of course, we are augmenting technologies, different generic generative AI technologies and using them as a layer, if you would like, within Enlighten in order to make it more accessible in terms of what I call consumable responses of the AI.
So the mix of the two together is the winning combination.
Tim Horan : And any color on how much better — how much of an improvement you’ve seen in the product? Or is it hard to say?
Barak Eilam : We see — the past 12 months, we’ve seen a dramatic improvement. And needless to say, the more deployment we have with customers, the more use cases we have. So I think we’re seeing what is typical from a new domain or technology that the early days of initial deployments and it’s actually more the initial deployment, the improvement is exponential. And we have customers using Enlighten in variety of scenarios and got to level of precisions that are beyond their expectations and are extremely seekable to what they did.
Operator: Our next question comes from the line of Meta Marshall with Morgan Stanley.
Meta Marshall : You mentioned predominantly consumption. But just wanted to get a sense of does that — does the pricing increase if there’s — as completion rates increase? Just how dynamic is that pricing? And do you feel like it’s still dynamic from your part and that you guys are still figuring it out just as it ramps maybe as the first question?
Barak Eilam : I apologize. We missed the first part of the question, but if I think I got it, the question was about the pricing, the consumption-based pricing of Enlighten and how stable it is and what we kind of see in terms of dynamics? Is that — that was the question?
Meta Marshall : Yes. Just is it are you guys still trying to figure out the pricing? And do you feel like there’s opportunities as completion rates increase to kind of increase the pricing that you’re seeing on a consumption basis?
Barak Eilam : So we have a pricing out with customers, and it is right now received very well by enterprises and customers. And the reason that I believe it is accepted by customers right now is because of the tremendous ROI it has. I’ll remind us again that a cost of a service representative in a typical CX environment is roughly $50,000 to $60,000 a year, and this is just the labor, not the technology. All the technology that’s supporting that agents usually are roughly $5,000 to $6,000. So 90% of the cost in CX still goes — or the investment still goes into labor. So when we replace or augment certain tasks or make those agents much more efficient or completely take them out of the equation, we tap into this significant labor [count].
Hence, that’s the way we look at it. And we can get 25% of those savings, it’s a tremendous uplift for our revenue moving forward. So right now, it is accepted by customers, and we see the potential. I believe there will be certain dynamics up and down. But we are in early days, and the land grab situation here is the most important thing.
Meta Marshall : Got it. And just as a follow-up question. Are people kind of taking the time or elongating the time to implementation to really train systems with kind of their own data or train kind of Enlighten with their own data? Or are they kind of leveraging your vertical specific expertise kind of out of the box initially?