전사 🔗
Russell: And and coming down to that, you know, say you see all these tool providers now building up their AI capabilities in this platform or that platform. what is the platform that really has a a philosophy that is overarching your whole organization? that's what I think is is is is is really hella
Caspar: Yeah.
Russell: highly relevant which that strategy you you go in that context and which providers really are are going down that path.
Caspar: Ladies and gentlemen, welcome to a new episode of the BPM three sixty podcast. My name's Casper, your host, and with me is as always, and I'm very happy that he as he still is, my good friend and co-host Russell. Russell, how are you, man? Where are you?
Russell: I'm at home, as you see, my winter garden. Yeah. And we and
Caspar: yeah y you you you went in a different place. You're trying to throw me off then.
Russell: we got the gardeners in the garden, so it's too loud there. So I I I I moved inside. Weather's fine anything I die. So yeah. So I I think it it it's coming down to a bit of a summertime, isn't it? so there's still there's
Caspar: It it it is, yeah.
Russell: a still lot on but I'm still every talk we have with customers is first of all, okay, I'll be on vacation then and then and
Caspar: Yeah.
Russell: What do we do about this project? So every year the same situation that you start the year off, then you have big high ambitions, and then by the time you get the budgets and the projects running, everybody's going,
Caspar: Yeah.
Russell: hmm, summer holiday. In your project plan, did you consider that we have a
Caspar: Yeah, did you consider that I'm I'm off for the next three weeks? Yeah, that's that well yeah, but we have this. Well that's your problem now. It's it's lovely.
Russell: Yeah, yeah. And and typically the situation is and that maybe also around process management as well. You you will have people will be there to cover, but it's just astonishing around projects, you need decisions. And and the
Caspar: True.
Russell: and and crude true delegation of decision power, that's when you see if that works or not in a in a summer holiday. Yeah, so we said,
Caspar: Yeah, yeah, yeah.
Russell: Yeah, yeah, yeah, I'm taking over. Yeah, yeah, yeah. But I don't don't I'm not gonna decide that. Yeah, we we're gonna have to wait ten.
Caspar: Yeah yeah.
Russell: Back, we have like a an alignment there. Which
Caspar: Yeah, yeah.
Russell: which I think with process management is a is an interesting topic to say, how do you really set up not just the owner and the expert, that you really establish sovereignty in taking decisions in a in a in a let's say in a situation where just somebody's off or not available? How often have we been in situations where the stakeholder was not just on holiday, he was business traveling to China to make sure the project's running?
Caspar: yeah, yeah. Yeah.
Russell: And the whole
Caspar: Yeah.
Russell: team back in in Germany is there going says, Yeah, well, he's not available, you know, it's a time zone combinate combined with workshop with customers. And and so building up a certain resilience around ownership for people to be able to take decisions on an operations level. Maybe that's
Caspar: Yeah. Yeah.
Russell: more the case that they they they they know they have to do that or else the the supply chain just stops. But also in a transformational perspective, I think this is a a big topic where every summer you have your test on how
Caspar: Yes.
Russell: how can you manage continuity in
Caspar: Yeah.
Russell: in in your projects during such phases where certain key players just are on the beach drinking their daikini.
Caspar: And and to a certain and but lesser extent the same happens around the Christmas holiday season and the new year, where where
Russell: Yeah, yeah.
Caspar: a lot of people also take some time off. Maybe not the Americans because they don't have that many days, but the Europeans still still do. but that's not the topic I had in mind for today, Russell. and I'm also not gonna talk about the World Cup. that's also for a different because the Germans and the Dutch are both out, so we don't have nothing to lose. we all know who's gonna win because it has been fixed that way apparently, but Let's not go into that rabbit hole. I would like to talk to you, Russell, about the the role of the process model, where it came from, and where it's going to. And I think that you you you you made a very cool statement in I before we started recording, so I'm going to repeat that. Process modeling is dead, long live process modeling. sort of. I mean, long live the process model.
Russell: Mm-hmm.
Caspar: That might be better. So we we need to talk about that. let me give the context and then I'm gonna I'm gonna set you free and I'm just gonna add some really annoying questions while you do that. So we all know where the process model came from. And and I I think for the people that might not know that, but I think ninety percent of our listeners know this. The process model initially was basically the deliverable of any BPM project. it was what you worked towards. When you've modeled all your processes, you were done. You had achieved the project. And then along the way, I think from the 90s into the early 2000s, it started changing, and people started to understand that your process model is something that's very useful, but it's used as input for bigger things. and I think right now, and this is the the I don't want to spoil everything, but I think right now we come to the conclusion that there is different ways to get to a process model. But the process model is essential if you want to bring context to any type of AI implementations. And I'll stop right there, but this is something we need to talk about, Russell. So g just give me your first your first shot.
Russell: I I think it it it's really well put in the especially as you said, like that the process model was something you as a deliverable
Load full transcript
that you work towards, yeah. And and now it's a process model is something we work with. Yeah. So and and
Caspar: Yes. Yeah.
Russell: and and you said that the of course people working with cha the models again and I think process mining is a good example where it has been understood that you to get the context you need also the process model.
Caspar: Mm-hmm.
Russell: I'm not going into the technical details if the BPMN was sufficiently modeled to be able to.
Caspar: No, no, no, that's not that not even interesting.
Russell: Essentially, it's the the actual content of the context that the model provides is also being something where more and more was what was understood. It's essential to actually be able to interpret the process mining data and the process and use the Process mining capabilities in a manner that you can make sense and take the right decisions. And also set a guideline for the transformation to say what are we heading at and how can we achieve alignment. So I've experienced it over the last years. Certainly, this working with the model has always been something that I believe has also been a strong asset for business transformations where. Coming from the let's model, do a blueprint, an RFI, an RFP, and let the system integrators just do whatever they've been they they they think they need to do. And then afterwards
Caspar: Yeah.
Russell: they go, we failed, how come? Because they weren't weren't working with the process models. So they weren't taking the they
Caspar: Yeah, that's a lot of double ups, yeah.
Russell: weren't taking the the blueprint along the course
Caspar: Yeah.
Russell: of their transformations. That's something I think fundamentally, this maturity on this level of understanding process mining. needs a model, a model context. A business transformation needs a model to actually structure its transformation along. Working with models has grown over the years, even before AI. And and that's and I would say, but then all of a sudden AI kicked in and and that's where two things changed. One of them, how do we get to a model and what does it have to look like? It doesn't have to just be pleasurable for the human being, yeah, and
Caspar: Yeah.
Russell: and let's say be stripped down, simplified, or put down in a way they said, that looks agreeable, but it's not really helpful. So, so it it it's a big shift to say it has to be able to answer all the questions that that we have in the context of our mining
Caspar: Mm-hmm.
Russell: analysis transformation projects. But I have AI that is able of let's say is agreeable with much more complex models. Yeah, so so that's where it it can manage them. And and that that
Caspar: Yeah. But that
Russell: is something that's really changed.
Caspar: but that's an it I mean you you you touch upon an interesting point here, because one of the conversations I have a lot with my customers is to what level of detail do you need to model? and on the one hand, if you if
Russell: Yeah.
Caspar: you model for let's say consumption, you want humans to understand the model, you don't need or you should not even go too deep because they get lost in details. However, if you want to model for usage in AI. You need a certain level of detail because the more detail you bring to AI, the better it understands its boundaries and guardrails. So that this
Russell: Exactly.
Caspar: is going to be quite a a conflict, I would say. Or maybe, and I just want to get your your your ideas on that, it could also be that we need to find a new way of documenting or capturing that information and simply prepare it in different ways and on different granularities for the use case for which you're going to use that information.
Russell: Exactly. And and that's the main point where I also say a big shift. we were limited due to the tools and we were limited due to say the the incapabilities of us human beings in in in modeling. So so what do we do to overcome that? And that that's the same with all models. You create models to be able to understand a complex world.
Caspar: Yes.
Russell: One of the key key aspects that in in process or in business process management is is one fundamental aspect we've always talked about as being process architecture. And how many
Caspar: okay. Yeah.
Russell: t even just the last couple of weeks, how many discussions said, is that level two or is that level three? And somebody says,
Caspar: Yeah. too too many.
Russell: that's level three, a week or level three, level two. And then you have to say, That's a process, level one, that's a process, level two, and you go, Hey, isn't a process a process? So What differs a level process level one from a level two? And isn't that just a structure? So you get into this this discussions where you said this is becoming absurd. Seriously absurd.
Caspar: Yeah, yeah, yeah.
Russell: But going back to what we did, so what we were doing is we were and I'm saying we, everybody in business process management was on this trip,
Caspar: Yes, yes. I get that.
Russell: yeah. So you you you build up your your hierarchies, and it's like, yes, you know, and and hierarchy is always good because that humans love hierarchy. Yeah, say the cupboard,
Caspar: Yes.
Russell: you open up, you have one cupboard for your summer clothes, you open it up, and you have it's
Caspar: It gives structure. Yeah, it gives structure.
Russell: a nice, beautiful structure. But what we were doing with the architecture as well is we were also looking at the layers, the levels, level one, two, three, four, and saying, what, who is this for? So if I think and I think that still makes sense. So you would you're looking at say, I need a certain level of aggregation and to be able to take certain decisions, and I'm not really interested in all the details underneath. And so we would
Caspar: Mm-hmm.
Russell: have like something like the Our thinking a lot of end-to-end scenarios. So you needed a domain with end-to-end scenarios, and
each scenario had a certain like objective, a description, and then again was decomposed into a sequence of processes, each process into a sequence of activities. And so we defined very distinguished levels and aggregation levels for dedicated target groups, but also for taking dedicated decisions and managing them on this level.
Caspar: Mm-hmm.
Russell: Is that still valid? I would s still believe yes, but it always going together now with the tools and the modeling, the act of modeling itself, it always
Caspar: Yeah, yeah.
Russell: was very cumbersome to say, okay, so I have to have a map and an object, and which conventions do I need on this level and how do I distinguish between level one and level two when I'm navigating that I can go a breadcrumb up to so I know which level I'm on. So there's a lot of I would say, to a certain extent, a lot of overhead in Defining the architecture, where you said, okay, let's just take a reference, you know, you you need to be able to
Caspar: Uh-huh.
Russell: work on that. then and but then you know, setting up this this this construct and then distinguishing on which level how much detail can you allow, or else it'll be too much work to maintain it, and it'll
Caspar: Mm-hmm.
Russell: be overboarding in the visualization. So you get tools like RS that have like dynamic filters, you can blend in and blend out things, you can do step views. Everybody had their own approach on in a tool, making it somehow manageable and usable for different user groups.
Caspar: Yeah, yeah.
Russell: And the whole business process management, let's say, community has been revolving around that. But ultimately, they always need this fixed layer, and you're always simplifying your user group. You're always having to because how many different roles and user groups and levels do you want to have? So let's just take three. That's already a lot of work. And
Caspar: Yeah. Yeah, yeah.
Russell: then an IT view and a business view. I think that's changed fundamentally. I think now, with AI and let's yeah, and and the tools we have, you can have in the background one, I would say, let's say let's say very Detailed model, I'm not gonna say an independent from levels, but you you have to have a certain like meta model behind it. And that
Caspar: Mm-hmm.
Russell: depending on your question, who on it has needs to see a certain set of this information, that you you will be able to say, that's a level two and a half I actually need for this decision taking. And
Caspar: Yeah, yeah.
Russell: I need it more from the view of compliance or more view from IT in the context of the applications. And that's the view, create me that. Show me
Caspar: Yeah.
Russell: the context, the model, but put a very
Caspar: Yeah.
Russell: specific filter on. And I don't want to be fixed just to these three or four layers where I say this is level one, you don't like it, then go to level two. You don't like it either, let's create a report. Or you know that'll be that's hard work. And and and that's where I think things are changing. So we can be much more complex in the models that we we build. They can be consistent because they have one common repository, not layer for layer that are then in in itself a lot of the times inconsistent. You see the flow of value processes, you see
Caspar: Yeah.
Russell: the flow of BPMNs, you said, that's not the same. Where's the process interface? How does that fit
Caspar: Yeah, yeah, exactly.
Russell: together? Yeah. And and so even and and that you can overcome if you had, you know, with with the help of AI, but also then the very specific, let's say, s snapshot or filter that you set on on on the repository. And and so the model has become more and more valuable because you can use it for many more use cases in a much more sophisticated manner. AI can help you maintain and update it and keep it consistent in on a on a next level without having loads and loads of consultants working on it, but you can also consume it better. And and that's that's a
Caspar: But but let me let me just let me ask a question, Russell. because what you're describing, I mean, it resonates really well, and I think that's what we're seeing in the market as well. I mean, what you're describing is instead of saying I'm going to create a repository with a lot of processes and potentially a lot of redundancy because I have different types of audiences to which that content needs to be directed, I'm actually going into one You can call it a graph or a context model where I store all of the information that's relevant and I connect it to each other. And then based on the use case, I simply construct at real time, I construct the view that you're looking for. And that that would mean that you would need to know in in instead of having to know this is the information that a certain person in the organization needs, somebody you're working for risk management needs a different view on on the information. Than somebody working in supply chain. So at the moment that you start defining the use cases, what you're basically doing is you're moving from a process-centric way of documenting your processes, you're moving into an object-centric way of documenting your processes in a single context model. And you basically tell the model, I'm a risk management person, and you can use AI for that. I'm a risk management person. This is my This is my area of expertise. So I I deal with the finance and and let's say finance and R and R D. Show me what's relevant. And it it will build up that information straight from that single context model, which by the way, then contains both the, let's say, the static documentation side of things, but also the more dynamic execution side of things that you can get from mining. It needs to belong into a single repository or a single
Russell: Yep.
Caspar: context model.
Russell: Yep, absolutely. And and that's where I think a lot of the and there's a lot of alignment happening
now on the let's say when these single repository tools now are are thinking about integration, you know, it's between a BPM tool and le you know, enterprise architecture tool, there's always like this discussion around the necessity for integration, but also b to Before you get down to the technicalities of the the actual the contextual alignment on the meta model that is necess necessary so both parties can talk to each other, even if they are using their own capabilities and their tools. So we and I think we've talked about that a lot. And it's just going to
Caspar: yeah.
Russell: show that more and more this is very unsatisfactory to a certain extent, especially in the age of AI. So And and coming down to that, you know, say you see all these tool providers now building up their AI capabilities in this platform or that platform. Yes, they're interested in making their platform relevant and keeping up their license sales, et cetera, et cetera. But from a real enterprise company view, you want to say, okay, so what is the platform that really has a a philosophy that is overarching your your your whole organization? It's not like they're locked into one specific let's say set of a repository or set of processes and and that's what I think is is is is is really hella
Caspar: Yeah.
Russell: highly relevant which that strategy you you go in that context and which providers really are are going down that path. And and here again, you know looking into the past and other tool providers. there's some out there that always had a very holistic view on the repository. And but they're But they they still haven't really understood I I think the value of of really let's say the modeling part of it is still so much inherently in their their genes that they they've not stepped out of that thinking
Caspar: Yeah.
Russell: of that that eighty percent of the functionality is around modeling rather than actually extending the capabilities of of let's say generating more comprehensive graphs and models and context and
Caspar: Mm-hmm.
Russell: and consuming them more more dynamically. And and I don't know if you remember in the time, so every time you had a certain specific question, and
Caspar: Mm-hmm.
Russell: this goes for data as well as for processes, you said, yeah, that's a very specific request. That's what I was saying about the the levels, if it's not covered. They said,
Caspar: Yeah.
Russell: you know what, we'll create a report. So we do a query,
Caspar: Yeah.
Russell: I do a little JavaScript or anything like that, collect the data, put it all together. And then spit it out in Excel. Look at it. Said, okay, let me work through that. Yeah. So
Caspar: Yeah, yeah. Been there, done that today.
Russell: exactly. And and the data side actually the same as well. Also here, you to be able to do a very big power BI reporting, you had to define your aggregation like logics over how you'd aggregate the data and then how you calculate KPIs and whatever. And very, very static like construct. As soon as anybody came and says, Yeah, but I want to have a look at a different like line of cost allocation or something like that. It's like, we'll have to write a report. So I think here again with AI, you can say reporting is dead, long live reports, because these
Caspar: Yeah.
Russell: questions they come down to a good prompt. And then you have a more or less if the comprehensive data model is is understood in its table joins and whatever. So like this
Caspar: Mm-hmm. Yeah, yeah.
Russell: whatever you need to to really understand the context of a database, then the the AI or agent can start looking after this question. I need to pull this, this, and this, I need to filter in this way, I need to aggregate and sum this, then present a pie graph, and here you are. And then you say, Hmm, Claude, that's rubbish because you forgot that and that is it. Okay, okay, goes up and then it gets it. Now, before that would have been a long discussion and then a report, and then you tell you you have to specify it really in detail and you have to yourself take care of the data model and say what you want to have aggregated. And a lot of these reporting activities in projects, they just took ages. The project was done and they still haven't created the report.
Caspar: Yeah, yeah, true. True.
Russell: And then it was already used once. I mean, how many SAP systems have you seen that they've done the analysis and they spit out a you re-documentation style? And you have a whole lot of
Caspar: yeah, I'd yeah, been there.
Russell: reports. And then there says, when was it last used? And how often was it used? So there's loads and loads of reports being created in the past. For even SAP, migration upgrades, whatever, that we use wants, maybe to consolidate the data, maybe to find some dead stuff.
Caspar: Yeah. Yeah. Yeah, yeah.
Russell: And and and a lot of effort has gone in there in the past. I mean, there's different le layers of reports. Yeah, so so you have like a financial
Caspar: Course.
Russell: consolidation report or monthly report versus an operative report. But a lot of this activity around there, I think in that case, it's more or less dead. Is data less relevant? Data models less relevant? No. more relevant than ever.
Caspar: Then absolutely true. But let me ask let me then add this and maybe ask you this as well. I mean, at the moment, you made you you made a remark that that really stuck by me. I mean, every tool vendor is is trying to get that AI angle in. Salesforce, SAP, all the CRM stuff, all the all the SaaS tools that you use as a company, they they bring in AI, right? But typically they focus on their own little part of the own little piece of the puzzle. Or little slice of the pie. So what you're really looking for in the end is an agnostic layer that sits either on top of that or simply takes in all of the information from the various tools that you use in your daily operation because you want the things to be
connected. It's not, I mean, as as beautiful as SAP is and as beautiful as Salesforce is, they don't represent the entirety of an organization. They always do a part of that. But if you want to use AI properly, you need to make sure that AI understands that if I do something here, it had has it has a consequence, maybe outside of this particular system in another system. And and so that means that even if you look at the SAP world, there is a lot of information there that you can use and you can put into a context model. And if you start mining it, but then and this goes back to the EA angle that you had, is like, yeah. This is the setup that we want as an organization. This is the capabilities we need. These are the applications that we need to realize those capabilities, right? And then I do mining and I do my improvements and I come to the conclusion: like, hey, we need to make a change here. That change
Russell: Mm-hmm.
Caspar: potentially has an impact on your capability overview or your application landscape. So you need to really. You need to string them together in your entire life cycle around continuous improvement. I still have no better word for it. to say, I know what I'm supposed to do, I know what I'm actually doing. Does that trigger a improvement action? Yes or no? All of that information that you need for that is in your context model. But for for your context model to be valid and and relevant. It needs to it needs to cross the boundaries of the individual systems, which a lot of companies still get into we're using the AI of Salesforce. Well, that's great, but that doesn't take into account what you do in SAP Ariba or what you do in God knows whatever system. so so
Russell: Absolutely. Yeah. And you will always need these separate systems. Or th well, mm,
Caspar: Yes.
Russell: yes, no, that then then
Caspar: Well, you need the separate systems to execute, right? I mean, they y they are your execution systems. They are not, by definition, your process management or process improvement or process intelligence systems.
Russell: Yes. So exact so so on that level of operations, I would say yes. And that's where say the companies have a PLM system, they have some LIMS system, they have an MES system, they have the so
Caspar: Yeah, EMS is and yeah.
Russell: so and and this this this orchestration of all this and and in the past a lot of this the the main knowledge on the context of let's say the PLM area, yeah, was also
Caspar: Mm-hmm. Yeah, yeah.
Russell: then very bound to the to the capabilities of the PLM tool. So you had like this this enclosed compartment that you could drill down, make your models, documentation, do your configuration and your transformation. But let's be honest, that it to say, is there a way of cutting a line, say, okay, Casper, I got it. Yeah, but aren't there areas that we can take out? Yeah? And then he said, Well, yes, and why would you want to take them out? Well, it simplifies life. You said Okay, okay, got it. Okay. So let's let's think about this one. Let's take out HR. You said, yeah, okay, HR. And and and then you go,
Caspar: Yeah, yeah, yeah. Let's do that.
Russell: yeah, could that work? And so if you're thinking about, say, outsourcing and then this, you said, yeah, maybe there's areas, but even there, or something like PLM. He says, no, but PLM, you know, if you scale up, if you want to do a sample, there's just so many, if you wanna do the cost calculation of a new product, you know,
Caspar: Yeah, yeah.
Russell: and then you want to do a product launch, you go into marketing and you're going into what is the demand for this product on the market that you can actually think about doing it or not. All of it, it doesn't matter where you go, there is, there is within, you cannot just say you're always simplifying and you always have to think about how much simplification you want to allow a allow when you say let's just let's do a hard cut here, just take that world and you do that world, do go into the lab. come back with the with a new new whatever product we can sell. And then
Caspar: Yeah. S something, yeah.
Russell: and then we we sell it and then we deliver it. And and you do notice how, and this is with with science down to everything, how collaboration and integration is it makes the cutting edge. Old organizations had their clear departments and everybody worked, but the cutting edge organizations are interlinked, they're collaborative, they're thinking b they're thinking of in their order to cash, they're thinking about the Products and services that they can develop for customers to be able to sell them in the future. And there's a closed loop on innovation there. So any organization that is really striving forward should not think, should think about ripping down these domains and thinking more and not saying, yes, but the system is in own application. It says, also you're limiting your capabilities to integrate and collaborate. Because a tool vendor says, just stick with that tool. and then
Caspar: Yeah, yeah, yeah. Yeah.
Russell: and then getting it into interfacing says, yeah, the interfaces technically is hard. Technically now, and that's another point where I think with the APIs, it there's a lot of precursors and parallel like developments. It's not just all AI. It's just I think that's just kicking off a a certain it's binding together these elements that we have. You know, and and that's where we talked to Alan. Yeah, there's been so many other developments like cloud cloud. and and data apis so the accessibility of you know connecting different systems and and looking at them jointly from a you know more higher level and and having more like a comprehensive view on it and then tools that can help us then much faster to to to get reports not do reporting you know and get get the models
Caspar: Yeah, yeah. Yeah yeah yeah yeah.
Russell: in place and building it together that is just that that's unleashing a lot of
potential is it being used yet Not to the full extent,
Caspar: Nope.
Russell: and and w we have to follow, but it's definitely it's more than a trend just to say
Caspar: Yeah.
Russell: this is it there's a lot of things happening in the last 10 years that are now converging to wow, let's make we can make something really great out.
Caspar: Yeah, I I I I agree. And I I think that even in the last six months, I would say that development is accelerating. you you you see left and right vendors starting their work on a unified repository where you where you can bring together physically
Russell: Mm-hmm.
Caspar: in one in one place the data that you get from mining and and and and discovery and the and the data that you have from your documentation and Coming back to your transformation remark at the very beginning, is it it's it's a triangulation comparison, basically, because you on one hand you have your your information of this is how you've been doing it. That's what mining gives you. You might have typically have information on this is the way you should be doing it, your document that should be processed. And then the third pillar in that triangle is this is how the vendor tells you you should do it, the best practice. And and If you only compare two of the three, you're always missing something.
Russell: Yes, good point.
Caspar: You're either missing this is what we told people to do, so you can't compare it to how they actually did it and learn from it. Or you you're missing the bet the vendor best practice, and then you run the risk that you're gonna configure or customize your your system too much, which in the case of SAP or any ERP system is also not good. so you you really want to be able to make that and and That is so complex that that is where you need AI for in the future to basically say, listen, here's your discovered data, there's your vendor best practice, there is the way that we that we were supposed to be doing it. What would you recommend us? What would be the best way forward? And
Russell: Mm.
Caspar: and that is why I say you should not only rely on the AI tools that come with the specialized tools for your execution. They should roll up into an overarching AI that basically can look into every angle of that. We cover every angle in this podcast. So in this case, that would be very relevant. You look at every angle of the information that you have and you come to a conclusion. Say, listen, I know that this might be the best solution if you only look at this particular tool. But if you take into account that it goes downstream into these particular tools, it's better to take a slight suboptimal view here. In order to be much more optimal, downstream the process. That is something you can only do if you have a single context model with a single AI sitting on top of that, branching out into the various. I'm not saying that those AI tools and operational tools are not good, but they should be, they should be monitored and coordinated through an agnostic AI layer that looks into everything rather than just one single part.
Russell: Mm. Yep. So that's I I I think that is the the the the actual that's the the the philosophy on on on which all vendors should actually look forward on. And and as I said, I th I think in the context of their their realm, definitely what what Jules planned there is is something in in their realms. They they want to provide a a complete holistic ecosystem for for for for businesses is is that
Caspar: But
Russell: that is that the decision an emp enterprise takes that's that's different take and there is a world beyond SAP which for people in SAP may not be so something they look at so much but even there a lot of their concepts go beyond just the SAP itself. But there's also
Caspar: Yeah, of course. But then
Russell: other vendors out there and I think these these vendors and I I think you may also have the one specific in mind that is is following along.
Caspar: Might be. No, but listen, let me let me let me add this.
Russell: No, no, no, no. I I'm I'm with you. But but I th I think that'll determine that this this phase now, that'll determine really the the real players that'll be of great relevance in the next couple of years. Yeah, so so there will always be niche and a little thing here and little things there. But the real
Caspar: My my last remark yeah, I I agree.
Russell: the the the the real and and it doesn't always have to be the big, big, big, you know, enterprise company, but
Caspar: No, no, no, of course not. No, no.
Russell: the the ones that tag on and and clearly so I I'll take just one step back. I remembered as a lot of BPM tools started offering SaaS solutions. And there were some that didn't do it at all or very late, or they had a lot of heritage that they had to carry over in that phase.
Caspar: Yeah. Yeah.
Russell: And there were others that just went cloud first, read rewrote their code, maybe not the database. I still see the schema that were very familiar.
Caspar: Yeah. True.
Russell: but they went like cloud first. GBTEC. And they did a really good job because it was
Caspar: Yeah, yeah, sure.
Russell: it really was you know, they they really did say, okay, this is the no paradigm, we need to get a software out that has bap, bap, bap, bop, but they did it, did it well. and and then others they did a half-half, and and and other ones just took a long time getting there. And they they were lagging behind for a long, long time before they really got onto that level that they could say we have a true ZAS offering. Now they They did manage to catch up the other providers and and I I don't think the impact and the speed with which you had to follow was so was so critical. I think now, if you look at it, there will always be niche players and small specialty or solutions that
Caspar: Of course, of course.
Russell: are really there. But but if you're looking at the real relevance, what is happening, I think it's you we
will see a lot of applications disappear. I'm I'm just talking about something like like
Caspar: No, I I I agree. So
Russell: Something like we we're on a level that certain applications become irrelevant because they don't they don't they they're not really needed anymore, like Photoshop. Something like that.
Caspar: No, no. I I get it.
Russell: I mean some people is gonna shout at me and say, Yeah, we we'll need Photoshop forever, maybe. but but there's a lot of in
Caspar: pro probably not, but I get the point.
Russell: that in that width of just doing some some basic operations around photos is just gonna be completely let's say automated and the tool itself had didn't have the relevance
Caspar: Yeah, of co of course.
Russell: it had ten years ago. In a business process management environment with the whole spectrum FOM2, yeah, that we always discuss, not just looking at the modeling part of it, talking about
Caspar: Yeah, absolutely.
Russell: the full stack. Each and every segment in this area, from the workflow automation, you know, low-code modeling up to the mining, sophisticated, whatever, it's it's all going to be determined by let's say understanding this core philosophy of of orchestration and the way
Caspar: Yeah.
Russell: context and the models are significantly being shifted. And it's not just the prompt to get an answer out of some static,
Caspar: Yeah.
Russell: like whatever model that's taken ages to create. It's it's it's it's it's a it's a complete different approach you have to inhale into your into your architecture of your application, capabilities
Caspar: Yeah.
Russell: and the organizations working with it as well.
Caspar: That that said, we we need to start wrapping this up, Russell, because I unfortunately have a bit of a hard stop in a couple of minutes. which is fine. but it was it was this was a very I I had a couple of points, but we we'll leave them for the next episode. I first wanna I first wanna see how how my holiday unfolds unfolds and then and then we'll
Russell: Yeah.
Caspar: get right back afterwards. so Russell, thanks very much. I I I think We did come to the conclusion that process modeling is that long live the process model. that's absolutely true. I think we see a lot of developments in that area, left and right, at multiple multiple vendors, multiple companies. it for me it's always interesting to see if if you are a a software vendor both in the execution space and in the process management space.
Russell: Yeah.
Caspar: I'm I'm I have never doubts on the good intentions of companies. I just want to I just want to know. How do you juggle having two hats and how do you keep your AI system as agnostic and and independent as possible? so that you you don't let's say imagine that you're a a big world soccer federation and you have the last semifinals or quarterfinals coming up, and then you assign a referee team, which belongs to the country of another country that's still in the tournament, which to me seems like that doesn't sound Really like an example of integrity, but I'll I'll let that one slide for now. I'm not a soccer expert, but for me, I know that that is something I think that was not possible until very recently, that referees could
Russell: No, no, they wouldn't.
Caspar: not be from countries that were still in the tournament. so let's see how France and Morocco fare by a completely Argentinian referee squad. It's gonna be very interesting
Russell: Yeah, yeah.
Caspar: to see that. But anyway, Russell, thanks very much, mate. Well and and to our listeners, thank you for sticking with us in this these this rambling about process modeling and context modeling and that kind of stuff. And we'll hope to catch you next time. Thank you very much. Bye bye.
Russell: Bye.