It’s like if you ask a 10,000 questions, does it come up with the same answer? Or does it start to hallucinate and kind of give some odd feedback to a client that — or to a customer that people don’t want necessarily going direct. And then two, around the security of their data and around who owns the data. Most of the generative AI kind of implementations and instances and everything like that, right now is you’re giving up unless you’re spending a vast sum of money, you’re giving up sort of that data to the generative AI cloud providers in order for them to kind of tune their models, make things better for everybody. And we just — when we talk to our clients, they are just not prepared to do that yet. Will that change? Possibly. But right now, they’re more focused on how do they own the data, how they own the learnings, how it aligned to their brand, how does it give predictable results, how is it secure?
And so, a lot of that focus is now saying, okay, well, let’s focus on proficiency of our staff delivering services because if there’s a hallucination or something else, then a human is there to make sure that it’s safeguarded. And then also the queries and back and forth, the human can handle a lot of the back and forth and really then kind of search and engineer the prompts for lack of a better term, to what the information is that, that customer is looking for much faster. And that’s what we’re seeing in our proof of concepts. That’s what we’re deploying right now. And we’re seeing the benefits as expected. We’re seeing faster resolution time for customers. We’re seeing higher CSAT scores because it has faster resolution. We’re seeing an easier experience for the adviser in order to get the information.
And because we’re able to get things done faster, the clients save some money from it because they have more capacity with less — well, they have more capacity for less money in effect is the way they look at it, if that makes sense.
Joseph Vafi: Yes. Fair enough. And then we’ve seen some of the other players, I guess, in the broader ecosystem, maybe especially also over in the IT side of things kind of announced like major AI investment initiatives of their own internally to build the practice up. Is that something that makes sense for Concentrix to do and kind of earmark real material budget for this stuff at this point? Or is it really a function of helping clients on their journey and maybe spending on investment there as it comes along?
Christopher Caldwell: Yes. Joe, that’s a great question. The reality is, a lot of press releases, I think, are somewhat misleading around where the spend is going and what it’s for. The reality is that, we are building up the practices pretty aggressively as we speak. It’s an extension of learnings we already have. The investments that we have put into Catalyst are significant, and that gives us the ability to scale sort of the proof of concepts much faster into production. And as a whole, as a company, we spend tens of millions of dollars on R&D in our own technology and development. That has been slowed down or stopped or anything like that. We continue to see that as critical investments for our future. And so, I think we’re at the right investment level and we’ll continue to ramp it up as we drive more and more into production and win more and more business around it.
Joseph Vafi: Fair enough. And then just one final one on the large Catalyst project that’s been — I guess is it fair to call it delayed and not canceled at this point? And then, is there any kind of view as to — if it’s just delayed, when it may start to ramp again? Thanks a lot guys.
