Celonis Celosphere 2025 | Rafael Domene, Cosentino

Clips
More from Celonis Celosphere 2025

Rafael Domene

Global CIO

Cosentino

play_circle_outline Savannah Peterson introduces the show and co-host Rob Strechay at Celosphere in Munich.

play_circle_outline Enhancing Business Efficiency: Strategies for AI Implementation and Automation Prioritization in the Growing Digital Workforce

play_circle_outline Importance of contextualizing AI and process intelligence for operational success highlighted.

play_circle_outline Bridging the Gap: Enhancing Human-Digital Communication through Celonis' Process Visibility and Tackling Siloed Data Challenges

Info
Transcript

Rafael Domene, Cosentino

Rafael Domene

Global CIO Cosentino

In this Celosphere 2025 segment, theCUBE’s Rob Strechay and Savannah Peterson sit down with Rafael Domene, global chief information officer of Cosentino Global S.L.U., to unpack how process intelligence is powering real outcomes in artificial intelligence. Domene explains how bringing core processes into Celonis – not just to observe, but to operate – created “control towers” that identify bottlenecks, prioritize fixes and surface automation opportunities. He details how Cosentino’s digital workforce has expanded from an initial autonomous engine built in sev... Read more

explore Keep Exploring

What event is taking place in Munich, Germany, and who are the hosts involved? add

What is the approach taken to manage and improve processes using Celonis? add

What are the challenges and considerations involved in implementing AI in large-scale operations outside of a controlled environment? add

What are the benefits of using Celonis for organizational visibility and connectivity? add

bolt Powered by CUBE AI

Rafael Domene, Cosentino

search

Savannah Peterson

>> Good morning, process-intelligence fans, and welcome back to Munich, Germany. We're here midway through day two of our live coverage of Celosphere. My name's Savannah Peterson, bringing you all the best and brightest with my favorite co-host, Rob Strechay. Rob, do you feel smarter than when you woke up today?

Rob Strechay

>> I absolutely do. I think learning from people that are in the trenches dealing with getting to AI being real is so fun and understanding how they're using process intelligence to get there just is great.

Savannah Peterson

>> It really is great. Speaking of great, it's wonderful to have Rafael back for the second year in a row. Welcome back to the show.

Rafael Domene

>> Thank you very much, guys, for having me here. It's a pleasure.

Savannah Peterson

>> It's always a joy. I was excited to see you on the lineup. Give us just a quick update on what's happened for you and the team over the last year.

Rafael Domene

>> Well, many, many things happened, all very, very good. Exceeded our expectations, because as I told you last year, we were doing a lot of things and it was very challenging because we were not completely sure how far we could go. So, as you know, we started our first autonomous engine we put in place early last year and after that-

Rob Strechay

>> Right. It only took you seven weeks to get there too, I remember that.

Rafael Domene

>> Yeah, exactly. Exactly. That is still working and it's doing amazingly well. And there are many, many now like that other one. We studied the processes we're covering and we start moving forward on using AI for more things than just AI agents to provide or automate certain functions within the processes. And what we're doing now is making use of AI also to provide more insights and more visibility on what's going on in the organization.

Savannah Peterson

>> So, talk to me, and I know you balanced your literal workforce as well as your digital workforce. How has that digital workforce expanded and evolved in the last few months?

Rafael Domene

>> As I said, a lot. It is growing very, very fast because-

Savannah Peterson

>> So, what does that look like exactly?

Rafael Domene

>> Okay. What we have done is the philosophy is we bring all our processes into Celonis. As I told you, Celonis is the one telling us what's going on, but our intention was not to Celonis as a way to understand what's going on. It's as a way to manage the processes and operate the processes. So, we build this, we call control towers, so that the person in charge running the process can see the process, can control it. But at the same time for Celonis for us is a very good way to understand where are the bottlenecks, the deficiencies, and the opportunities for automation. So, we have done is finding all those opportunities, balancing which ones are highest priority. You're cannot do everything in one moment. And we put in priority on which ones we can automate and we can make the process seamless and working more efficiently. So, we're starting identifying in which part of the process is require an agent and we were installing them. In the same process we took, there's around 17, but in total they will be close to 100 now.

Rob Strechay

>> Wow.

Savannah Peterson

>> Wow.

Rob Strechay

>> That's impressive. I think part of it was also that you were talking about putting AI in the right context last year. Because I think one of the best things about Celosphere is that it's learning from other customers. So, help those who didn't make the trip here, per se, understand what does it mean putting AI in the right context, and how does process intelligence tie in with that?

Rafael Domene

>> Well, as I said before, it is critical for us. I mean, we've been building agents also outside Celonis, and it is tough. Because we are a company, we have massive operations, massive production, massive logistics. And building things with AI is relatively easy in a lab. But when you want to put that in production, you have to take into consideration other things. So, the safest way and the most productive way is doing it within the safeness of Celonis and the process intelligence because you have all the environment and you know what's going on. And as we were talking last year, this is us bringing a new person, a digital person, into your processes with certain background, but you trained them on other things and you have to monitor them. So, it's not about giving them the right context, it's not about giving them the right information so they can do the right thing, but also understanding how they're performing. Because as a normal human being, you can make mistakes. There's the things you haven't thought about and you need to know how to proceed with certain things. So, this is something that we have a lot of confidence. It's not only easier to build things there, but it's a very, very good way to control them, to understand them, to monitor them. So, I would say for us it's key, when you want to operationalize some automations, AI and not AI, you have to do it with the process intelligence.

Savannah Peterson

>> Absolutely. So, when we think about communication, obviously critical to any business, both human communication as well as digital workforce or agentic communication. How are you optimizing, enhancing the digital communication between the agents within your organization?

Rafael Domene

>> Well, I mean the good thing is Celonis provide you this layer that connects everything and provide visibility about everything. And that is not only for agents but the entire organization, humans and digital agents. Because the main problem you usually have is organizations work in like islands or silos. And you can have the best data in the world, but it's in a silo mode. And also, usually it's historical information. So, what happened yesterday, what happened last month, et cetera. So, it's really difficult to connect the things and connect the... Because normally, we don't work in an way you are doing something for someone else or someone is doing things that you need to continue doing. There's a process. So, having the process, linking all those silos and providing that connected information is key to have this communication. And on the technical side, I know you were asking them from that perspective or from the technical side. From the technical side, we're using typical standard like MCP or agent-to-agent communication to the... We have agents outside Celonis. We're working with Microsoft. Microsoft is our main partner to go on automation and we were very glad when Microsoft and Celonis signed the partnership. That was a blessing for us and make our life much easier. But we were doing some automations outside and many automations inside Celonis. And for us, was critical to have them talking and communicating. So, you have agents inside Celonis can be used outside Celonis, and you have agents outside Celonis that are used inside Celonis. So, it was not that easy because standards are changing as we speak. So, both on the Microsoft side, on the Celonis side, you have to be trying to catch up with those things and have something stable. But we have a lab, so we can play with many things and then we move into production things when we are completely safe and stable.

Rob Strechay

>> So, I mean you mentioned it and I think AI and people getting to agents, so many people, there's a lot of fear about agents because they see the numbers of failures that are out there. Even Alex was talking about only 11% are seeing value out of their agents. You guys have seem to really crack the code on doing agents right. Has that really been by understanding the process first and what your end goal is and having a metered way to go through that? Talk to how you've been able to prioritize that and get to it.

Rafael Domene

>> Last year when we started... Well, before last year, it was three years ago, I was very clear that we have to take advantage of gen AI. And the thing is that I was also clear that to make it happen, you have to have the right information, the right environment. That's why I went to Celonis and said, "Guys, I mean you are the only ones that can provide me this so far, as far as I know. I don't know if you are willing to do this." And the teams, Basti and Carsten said, "Yeah, yeah, absolutely. This is our vision. We're going for that." So, we start working like that because, for me, it's a must. I mean, working outside, as I said before, outside Celonis, outside process intelligence, outside that information is a very, very tough world. I mean, I'm not saying it cannot work. We have agents working outside, but within the organizations and production, you have to have them back up by all that process information. Otherwise, it's nearly impossible. It doesn't make sense.

Savannah Peterson

>> What would you advise companies, organizations that are about to deploy Celonis and really start to understand their process intelligence? What would your advice be to them to navigate what can be a bit of a complex and sometimes a little messy at the beginning process?

Rafael Domene

>> Yeah, you have to start small. Start small till you understand very well... And it's not only about the technical part of it's also the organization operations. The whole organization has to be evolved. This is not something about technology, it's about changing the way we manage things. And the good thing about this is now people in operations, they are now understanding and they can see things they couldn't be seen before. Once you start seeing those things, understanding all the benefits this can bring you, things start popping up and when we say you can look under the carpet and say, "Oh wow, there's a lot of opportunity here." And then, you can decide how fast or how slow you want to go depending on the organization. Because there's also one of the big challenges is not about implementing this, it's also to manage the change within the organization. And there are things that are really amazing and that takes some time to have the organization getting ready for those changes. So, that's the main thing for .

Rob Strechay

>> Just bridging off of that, I think, again, and change is one of those things is hard. I think rebuilding processes is hard and rethinking them. But when you look at it, what has the effect of AI been on COSENTINO from a day-to-day operations perspective?

Rafael Domene

>> Well, a lot because whatever we do has to have a very, very practical impact. It has to impact P&L. I mean it's critical. You have to be able to improve revenue. We're doing things on the commercial side for that, on the commercial and sales process for that. For operations, you have to improve mainly cost, improve performance, increase performance. And everything we do is discussed before with the business. So, business is fully involved there saying, "Okay, what are the main problems? What are the main issues? What are the main pain points? What do we want to improve?" And based on that, we go there, focus on how to improve those things. So, my point of view, that's the way to go. If you do that, then it's relatively easy because everyone understand all the opportunities and then you can implement them. We have implemented this in two major processes and we have the full order-to-cash process. And there was a specific area within the organization where we were having some issues because of several manual steps in the middle. And it was so difficult to find out where were those opportunities. With Celonis, we could find them very easily. And what we've been doing for last year is automating step-by-step in that process. Wherever there was an opportunity, we doubled service level, which was something very, very important. And the savings are amazing. So, every area, there's 15 different areas involved in that process and everyone is so happy, so excited about that because they can see improvements on a daily basis. It's really cool. It's really cool.

Savannah Peterson

>> It's really cool. I can tell just from your energy how impactful that is. So, how does all this work that you've done internally on your processes help benefit your end customer?

Rafael Domene

>> Yeah, a lot. As I said, the main target here was, okay, our service level can be improved. There was a side business where we have a lot of manual steps, as I said, and whenever you have manual steps, that impacts the process because we humans are not perfect and we have to be trained-

Savannah Peterson

>> We're not?

Rafael Domene

>> No, not always. Not always. Not always. Some of us, we're not perfect. So, you have to be trained and things happen. So, when you have some manual processes in your process, that may, in fact, impact the service level. So, we said, "Okay, how can we improve?" The whole target was to improve service level for our customers. We have customers that any delay is really important for them. Imagine you're building, you're making a construction and you're going to have all the fascia that is going to be from COSENTINO. They asked for permissions to the city to put the cranes, to put the scaffolding, to bring teams there. And we're going to go and say, "Oh, sorry, we are late. It is going to be one week late. We'll bring the material." It's a big, big mess, a big problem for them. So, being on time is really key quality-wise. Being with the right quality and the right time, it's very, very important. So, the organization said, okay, this is a high priority. Let's see how far we can go. And that's what was one of the first one. We started end of last year, but mid this year it was done. And what was very nice about this one because the problem was serious and then the said, "Okay, what about trying to solve it with the traditional backend tools we have? Because we have these backend tools for the last 25 years." And the said, "Okay, let's try to see if we can solve it." The timeline and the time necessary to solve the problem was around 15 months. So, the business said, "No, that is too long, guys. You need to do it faster. Can you do it faster because we cannot wait one year and a half to solve this problem?" So, the team started thinking about, we were working a lot with our processes. And they said, "Okay, why don't we try to do this with Celonis?" And they said, "If we do it with Celonis, we think we could do this in three months." So, we went from 15 months. It was not three months. At the end, it was three months and a half, so it was close to four months.

Savannah Peterson

>> Geez.

Rafael Domene

>> But we did things that we could have done with the backend infrastructure. So, it was really amazing. It was all for customer service, improving service level and the time it took from the Celonis point of view and from the internal users, because at the end of the conversations, you're not going to stop the business to do these things. They need to keep on running. So, you cannot ask for a lot of time, a lot of dedication to these things. So, if you do it in a shorter period of time with less involvement from their site, in terms of detracting time from their daily job, it's good. So, there was a very, very good project. All was initially for service level, but then the improvements in the organization went farther than that.

Savannah Peterson

>> That's exciting. That's a really significant increase in speed that impacts a lot of different things. Two final questions for you. You I believe are only guest from Madrid on the show this week. And I'm very curious because I've noticed a lot, Rob, I'm sure you have too, as we travel around the world, it shows the conversation around AI is very different in different regions. So, what's the buzz on the streets in Madrid?

Rafael Domene

>> Well, Madrid is booming now. It's really amazing. It's a place that you have to visit as soon as you can.

Savannah Peterson

>> Twist my arm.

Rafael Domene

>> Okay. Yeah, sure. In Spain in general, everyone is... I think that everywhere in the world, everyone excited about how far you can go with AI. We are trying to push internally and to our colleagues... We have CIOs associations and groups where we discuss a lot these things. And everyone is understanding that the potential of AI is there and traditional AI has been there for a long, long time. Gen AI can bring a lot of things, but you have to be careful and things are not that easy. The potential is amazing. With the current state of technology, we can do a lot of things. Doesn't matter if they evolve or not. With the current state, we can do a lot of things, but you have to do it the safe way and in an enterprise way. I mean, the thing is to make things enterprise class is not that easy and you have to have the right tools. And what we are pitching is you have to know your processes and have the right information, which is not just data because we discuss this a lot. And normal data, everyone said, "Oh, I have a lot of data so I could use it and this is key for AI." Not really. It's not fully true. Data is absolutely important for organization, but you have to have process data, process information, otherwise those are silos that are not really helping what you need to do.

Savannah Peterson

>> Absolutely. Couldn't agree more. And it's fun to hear that it's so buzzy, really. I mean either of us would need an excuse to go to Madrid, but that sounds amazing-

Rob Strechay

>> I love Madrid. I've been before a couple times and it's way better than New England. There you go. There's actually really good fish there too. I have a great fish restaurant that I love going to when I'm there.

Savannah Peterson

>> Oh, gosh. Yeah. The food in Spain really doesn't leave much to be desired. It's fantastic.

Rob Strechay

>> Yes, I was going to say it's fantastic.

Savannah Peterson

>> Final question for you, Rafael. When we're hanging out at Celosphere 2026, because you're now a regular with us on the show, for this show specifically, what do you hope to be able to say then that you can't say today?

Rafael Domene

>> We are pushing the envelope a lot together. Celonis is a great, great partner to us because we both agree that we are going into unexplored territory for certain things. And one of the things we're working during this year, and it's already working, but we want to be sure we can take this massive... It's about using all the information you have in your processes to answer questions you couldn't answer before. We discovered this year, I call it the time travel machine, and everyone laughs when we say that, but it's true. With Celonis we can travel into the future and then go back traveling to the past and understand how things... And you say, "Okay, that is really tricky." No, no, no. It's not that easy to go and be able to go in organization and say, "Based on the information I have now and the knowledge of my processes, I like to know what is going to happen in the future." That is really important for organizations. And not only that, because when you know the future, if there's any problem, you can see any problem there. Now, you can go back to the current state or to the past and solve that problem before it becomes a real thing. And that is something we are already doing now. And something also very important is when you have an issue, it's important. There's something you didn't expect to be able to understand why it happened. So, we are using now gen AI to go and ask , "Hey, how is the business going on?" It will tell you, "Oh, how is an order? What is about this order?" The guy go and said, "What about this order?" And the system is telling them, "Well, the order is okay, after this tender, but you may have some issues in the future. This is the steps that has gone through and the following steps. In this one, you may have an issue and you should be working on that now." Things like that, which is really amazing and we're putting a lot of interest that next year we will be able to show you that, I hope so.

Savannah Peterson

>> Can't wait to see it, Rafael. Thank you so much for joining us again.

Rafael Domene

>> Pleasure, like always.

Rob Strechay

>> Thank you.

Rafael Domene

>> Thank you so much.

Rob Strechay

>> Yeah, thanks so much.

Savannah Peterson

>> Yes, cheers. And thank you, Rob. And thank all of you for tuning in, what is a really fantastic week here in Bavaria, in Munich, Germany. We're at Celosphere. My name's Savannah Peterson. You're watching theCUBE, the leading source for enterprise tech news.