Celonis Celosphere 2025 | Prof. Wil van der Aalst, Celonis

Clips
More from Celonis Celosphere 2025

Prof. Wil van der Aalst

Chief Scientist

Celonis

play_circle_outline Will: The Godfather of Process Mining Shares Insights on AI and the Rise of Object-Centric Process Mining Post-ChatGPT

play_circle_outline Mercedes-Benz and Unipad as examples of sophisticated AI applications in supply chains.

play_circle_outline Optimizing Core Processes: How Lufthansa Leverages Process Mining to Improve On-Time Performance and Reduce Airport Delays

play_circle_outline Process mining serves as a layer above existing systems to improve operations.

Info
Transcript

Prof. Wil van der Aalst, Celonis

Prof. Wil van der Aalst

Chief Scientist Celonis

In this Celosphere 2025 interview, Prof. Wil van der Aalst — chief scientist at Celonis and widely known as the “godfather of process mining” — joins theCUBE’s Rob Strechay and Savannah Peterson to unpack how object-centric process mining (OCPM) has moved mainstream and why there’s “no AI without PI.” He shares real customer proofs – Mercedes-Benz improving supply chain reliability with AI-powered predictions, Unipad optimizing facility maintenance, and Lufthansa reducing airport delays – showing how Process Intelligence, as a layer above scattered systems, d... Read more

explore Keep Exploring

What updates did Will provide regarding the current developments in AI and its application in object-centric process mining? add

What are some real use cases of AI and process mining in supply chains and business operations? add

What are some examples of how companies are using process mining to improve their core business operations? add

What is the role of process mining in relation to existing systems within an organization? add

bolt Powered by CUBE AI

Prof. Wil van der Aalst, Celonis

search

Savannah Peterson

>> Good morning nerd fam and welcome back to Munich, Germany. We're here midway through day two of our two days of live coverage on theCUBE at Cellosphere. My name is Savannah Peterson, bringing you all the details and process optimization and ROAI, with the fabulous Rob Strechay. Rob, I am particularly thrilled about our next guest, because he's a legend.

Rob Strechay

>> Yes, a legend and a Cube alum that's back again this year. Another one that we can dive real deep into what's going on in process intelligence.

Savannah Peterson

>> Without further ado, the godfather of process mining, Will, welcome back to the show.

Prof. Wil van der Aalst

>> Happy to be here. It's always a pleasure to talk to you guys.

Savannah Peterson

>> It is an absolute joy. So lots has happened over the years since we saw you last time, this time last year. Give us a little bit of an update on what you've been seeing.

Prof. Wil van der Aalst

>> Yeah, so I think AI is undeniable, right? I think many people post ChatGPT, everybody's talking about it. And what I find super interesting is that if I look at something like object-centric process mining, which used to be something novel and companies were not sure when and how to apply it, I think now the landscape completely changed. Everybody has bought into it and see that this is the way to go. And what you also see is that now people do not just talk about AI, but also actually apply it. And I think that's the biggest innovation, that both object-centric process mining and AI is not just talking, it's actually now the companies are doing it.

Savannah Peterson

>> So-

Rob Strechay

>> I...

Savannah Peterson

>> Go ahead.

Rob Strechay

>> Yeah, I was going to say, I mean, just bridging off of that, I mean, in talking to customers, what are some of the more sophisticated applications that you've seen with these customers where they're not just doing one task here and automating that, but looking at more how they rethink and redo their operations?

Prof. Wil van der Aalst

>> Yeah, so if you look at also customers that were on the main stage, like Mercedes-Benz, in supply chains it's very important that things are reliable and that you have an idea when something is going to happen. And they're both object-centric process mining, and the ability to use AI to make more reliable predictions, I think is a clear case of something that you could not do before. We also had Unipad, for example, on stage. I think they are using the AI part mostly to optimize the maintenance of the facilities that they're running. So I think these are use cases that are very, very real in general. Me personally, I always like use cases that are a bit closer to the core processes. Of course, many customers start with things like purchase to pay, order to cash, accounts receivable, accounts payable, and all of these standard processes where it's very easy to get started, because many people have done it before and it's clear how to approach it.

Savannah Peterson

>> Bit of a blueprint. Yeah.

Prof. Wil van der Aalst

>> So the threshold to get started with that is very low, but it's often not the company's core business, right? So BMW needs to produce cars rather than sending bills in the best way, right? And I think these are the interesting applications that I see. And so for me still, one of my personal favorites is, for example, Lufthansa, that is using process mining to reduce delays at airports. And these are things that are very close to the core processes of these organizations, and that's where in the end you can also have the biggest competitive advantage by applying it.

Rob Strechay

>> And close to our heart as customers of Lufthansa and .

Savannah Peterson

>> Well, yeah, I was going to say, I think one of the things, and I bet you experienced this too, we're all kind of facilitators of new technology. We're the types who help make it feel less scary and dispel myths and really normalize new tech for the greater population, not just for the nerds like us, but for our families and for folks. And when we're talking about something like Lufthansa, every person who has flown on an aircraft has probably experienced a delay at some point in their travel journey. And to be able to change that has such a positive impact, not just on the P&L or the bottom line, but on that true customer experience. Can you give us any other examples where you're seeing process intelligence really driving either science or the human experience in a similar fashion?

Prof. Wil van der Aalst

>> Yeah, so perhaps an interesting anecdote is that the city where I live, the city of Aachen here in Germany, I've been trying to convince the city to apply process mining already for a very long time.

Savannah Peterson

>> Shocking.

Prof. Wil van der Aalst

>> Not very successful, right? I tried to convince them, but then you were asking about reaching out to the actual people that experience these things. And then I was invited to give a presentation at a large event in the city where there were many citizens. And when I looked at the program, I saw that the mayor was just before me, so I think let's use this opportunity. So then defense started, the mayor talked about how great the city was, et cetera, et cetera. And then my first slide that I presented was a slide of ChatGPT saying, "I'm the mayor of the city of Aachen, and I have these and these problems." Right?

Savannah Peterson

>> Clever.

Prof. Wil van der Aalst

>> And then ChatGPT was, I don't know, generating pages of nonsense, general knowledge, without using any of the data available about how things in the city were running. And then I explained, okay, AI can help with these things, but you should first get the basics right. You should first get what are the processes, where are the actual problems, really evidence based. And then it was very funny, then I was done with my talk and it resonated very well. And then afterwards the citizens could ask questions related to that. And basically everybody stood up with the answer and they were all asking questions to the mayor, basically asking why she was not using process mining. And then you hear of all these experiences that all of us have where, I don't know, you have an application for something relatively simple and it takes half a year, or you get something digital, but you have to make a signature on paper and then send it back, et cetera. So it was clearly, I think everybody in the room was convinced this is the way to go. And I think these are the nice examples where people really experience process-related problems. And I think it's our goal to convince people, even though sometimes it's unpleasant, right? You need to look at reality and face it. That is the only way to go.

Savannah Peterson

>> So did you make some headway with the mayor? Do you think it's going to happen?

Prof. Wil van der Aalst

>> Multiple appointments in my calendar, now.

Savannah Peterson

>> Yes, I love to hear it.

Prof. Wil van der Aalst

>> That worked out, yeah.

Savannah Peterson

>> I love when a plan comes together. Well done, that was a clever approach.

Prof. Wil van der Aalst

>> Yeah, yeah, yeah, yeah, yeah.

Savannah Peterson

>> It was a very clever approach.

Prof. Wil van der Aalst

>> Yeah.

Rob Strechay

>> No, I mean, I-

Prof. Wil van der Aalst

>> In the same way I tricked my own university into doing it.

Rob Strechay

>> Always good when you can bring it back as an alumnus.

Prof. Wil van der Aalst

>> I know, but that's the point. We always think of the tip of the iceberg, the very sophisticated processes, but the huge potential is in let's say lots of organizations where really the basics have so much improvement potential, but people just need to get started.

Rob Strechay

>> I mean, I think you hit on this a little bit, which is people trying to get started, and where do they start, I think the addition with the GA of data core and some of that stuff, but being open, data-driven, and really composable, has been a big theme this week.

Prof. Wil van der Aalst

>> Yeah yeah.

Rob Strechay

>> When you look at this from a scientific perspective, how is that helping to really accelerate data mining and data modeling and AI orchestration? Because to me, we see the silos at every organization we talk to.

Prof. Wil van der Aalst

>> Yeah, so the way that I think of process mining is that process mining is a layer on top of existing systems. So any organization of some size is using multiple systems, things are very scattered, and process mining and process intelligence in the layer on top of all of these systems bringing the data together. So our goal is not to replace existing systems, because if you start trying to do that, then you get into super expensive IT projects that are very likely to fail if you don't do it properly. So it's this layer on top of things where the goal is to improve end-to-end processes rather than not an individual step in the process. And I think you need to do that, and of course that requires effort. The data is scattered over these various systems, so you need to do that. And instead of trying to replace existing systems, you see where the main pain points are. And for every pain point that you see, you need to think, "Okay, how can I now solve this?" And in some cases you need to change the organization, but in other cases it may be as simple as putting an agent there that whenever this thing happens, does something automatically in such a way that it does not exist anymore. So it's a very, let's say, flexible approach. Of course, how to get started. So many companies start with the proof of value project, and often it's amazing that just in a couple of weeks for a specific process, we can show, okay, if you're now using process mining, you can really improve it a lot. But that's only part of the story, right? Because it's typically focused on one process. I think organizations that really want to be successful in the long run, they need to approach it in a much more holistic way. You need not to focus on one process, you need to focus on the whole. And I think this is something where organizations need to invest in, and that's why it's super important that at Cellosphere there are also many decision makers that are able to make these types of decisions to do that. I've experienced it many times that in an organization, people are enthusiastic about process intelligence and process mining, and if it starts bottom up, that's a difficult story, right? Because you create transparency, you show things that are not always-

Rob Strechay

>> A little bit messy.

Prof. Wil van der Aalst

>> A little bit messy. So it's very important to have support that you can do this top down, which is good for the whole company, not just somebody's job, somebody in the pyramid.

Savannah Peterson

>> Right. Somebody who's progressive, risking their career to be driving forward this type of innovation.

Prof. Wil van der Aalst

>> Yeah. Yeah.

Savannah Peterson

>> You mentioned the pain points, a lot of Celonis covers pretty much every single vertical of business. Is there a common theme between some of the pain points you're able to solve earlier on for these companies, or is it very different depending on the vertical or the industry?

Prof. Wil van der Aalst

>> Yeah, so I think on the one hand, there are standard processes that I talked about before that exist in almost any larger company. Right? So that is uniform, that's the same everywhere, and that's how you can get started.

Savannah Peterson

>> Everyone's sending out invoices, or...

Prof. Wil van der Aalst

>> Whatever, but if you then go to the areas that are really specific for a domain, I think there is no recipe, right? You need to really create the models needed to get the data in order to do these things. And so, I don't know, reducing delays with flights requires a completely different approach than trying to optimize patient care ways in the hospital, right? That's completely different.

Rob Strechay

>> Yeah, I think again, tying it all together, and I think like you said, you have to start somewhere, but I always looked at it as crawl, walk, run, with this, and I did BPM way back in the '90s and with spreadsheets and Gantt charts. And this is so much better to me, and I think to your point, I think one of the things that we, and one of the common themes that's been throughout a lot of the customers who've been on, is using things like the process intelligence graph that was announced last year, and now the excitement with things like MCP and things of that nature. Do you see that being able to be part of that, how you can be part of that composable and share the data a little easier? Because MCP, again, agent to agent technology coming as well with A2A, and certain... Constantino was talking about that just a minute ago. How do you see this all progressing over the next year?

Prof. Wil van der Aalst

>> Yeah, so as I said, like with process mining and process intelligence, we want to be this layer throughout the whole company where we collect structured information, and that's the goal. On top of it, there are all these interactions with all the different agents. And there, to be very honest, I think the landscape is quite heterogeneous today. There are, as I said, some things you can simply address by letting a person do something different, other things require you are using sophisticated agents, but also don't forget about technologies like RPA that are also still there, et cetera, et cetera. So I think the landscape is very heterogeneous and therefore it is important to have this abstraction layer where you, at the organizational level, you have everything. And I think now people are, let's say, super excited about agents and stuff like that. I often try to tune down these expectations a bit in the sense that, yeah, this is another, let's say solution that you can use that will give benefits in the short term, in some cases very easy to apply. But let's be completely honest. If in the end an agent can do it, you could also have, let's say, programmed it in the way that it would work directly without the Band-Aid type of thing. So I think with new technologies lots of things are possible to do that very quickly, but we also have to realize these things are probably temporary. But getting this layer in your whole organization that everything is well organized is not temporary. That will remain valid also when trends and techniques change,

Savannah Peterson

>> It's a very good point, and it's nice to know that that effort, that investment, that commitment up front, it sticks with you, to your point. You're not going to have to retool shortly or have to redo something. You'll be able to consistently iterate and optimize and innovate. I got one final question for you, Will. Lot's changed in the last year, like we talked about to kick off the segment. What do you hope to be able to say at Cellosphere 2026 that you can't yet say today?

Prof. Wil van der Aalst

>> I think we should just continue what we are doing. I think we are on the right track. Many things are gradual. So next to my job as chief scientist, I'm also a professor, and many people ask me the whole day, "Will, what is the next big trend?" et cetera, et cetera. And I say, it's a bit like if I was a professor trying to cure cancer or something like that, you also would not ask that person, "What are I going to do next year?" Right? So the goals remain the same and there are these innovations that we see each year. And as I said, I'm already very happy, let's say how OCPM has now become mainstream. And I think it's very important that that continues. That will provide the basis for AI. I think also many organizations will be disappointed. I think Alex, on the main stage talked that like in some studies, let's say only 10% of AI projects succeed, let's try to increase that percentage a bit. But I think most of these things fail because people, as you just said, want to run before they can actually crawl. Let's show this very transparently and very clearly, and then next year there will be many more success stories.

Savannah Peterson

>> I love that. Well, we hope to tell all of those success stories. I hope that you're able to tell us a success story from your town and from your mayor as well with us next year. So I'll answer that one a little bit for you.

Prof. Wil van der Aalst

>> Let's see, let's see. Yeah.

Savannah Peterson

>> Yeah, yeah, yeah, let's see. I look forward to having the conversation, Will, thank you so much.

Prof. Wil van der Aalst

>> Yeah. Yeah. Thank you very much.

Rob Strechay

>> Thank you.

Savannah Peterson

>> And thank you Rob, and thank all of you for tuning in to our two days of live coverage here at Cellosphere in Munich, Germany. My name's Savannah Peterson, you're watching theCUBE, the leading source for enterprise tech news.