Max Aberman has been with ENERVIE ever since 2020 and currently he is leading the Process Mining initiative at ENERVIE through his expertise as a Digitalization Expert.
Episode 37
Process Mining in The Energy Sector With Max Aberman and Jannis Nacke, ENERVIE Group
Our special guest
Max Aberman
Our special guest
Jannis Nacke
Jannis Nacke has been with ENERVIE since 2021 and currently he is implementing process mining solutions at ENERVIE.
featuring
Max Aberman
Process Mining Expert
featuring
Jannis Nacke
Process Mining Expert
Episode Content
Do you know what happens behind the curtains of process mining implementations in the energy sector? Max Aberman and Jannis Nacke from ENERVIE Group sure do. In this episode they talk about process mining in the energy sector, industry specific processes and how process mining assists the different departments within the company. Maybe they also talk about how amazing Processand is as a company, who knows?
Transcript
00:00
Patrick
Knock, knock.
00:01
Jakub:
Who's there?
00:02
Patrick
It's The Mining Your Business podcast with yet another episode. A show all about Process Mining, data science, and advanced business analytics. Jakub, how are we doing today?
00:10
Jakub:
We are doing very well, Patrick. Thank you.
00:12
Patrick
Max Aberman and Jannis Nacke from ENERVIE Group join us on the podcast today to talk about how Process Mining is helping them solve critical problems in the energy sector. Maybe they will also mention how good we are as a company. Let's get right into it.
00:32
Jakub:
What time is it? It's Mining Your Business time. I hope you're excited because we sure are. Slowly, slowly, we are reaching another milestone - forty released episodes and we are starting to eye that elusive fifty-episode mark and, you know, things are going very well in our end, so we are very excited about that. And I was thinking, how is it even possible that in such a niche fields, niche topic, such process mining is we can produce so much stuff. And then I had it and it's not really us, but it's you guys, is the process mining community and people such as guests of today's episode are making it happen for us that first and foremost, we always have some intriguing topics to discuss, but also because these people just keep entertaining and educating us in all things process mining. And there's just so many stories and so many people we could invite and you know, that just keeps us motivated and keeps us wanting to keep releasing these new episodes. So really good stuff. And that being said, please welcome a pair of Process mining experts, Max Aberman and Jannis Nacke. Guys, welcome to our show. We are excited to have you.
01:42
Max Aberman:
Thank you very much.
01:43
Jannis Nacke:
We are glad to be here.
01:46
Jakub:
Now, this is actually exciting because not only you guys work together for the same employer in ENERVIE and in the same field of process mining, but you also seem to be, well, almost the exact same age and you actually studied together at Fachhochschule Südwestfalen. How did that happen?
02:06
Max Aberman:
Well, you could call it University of Applied Sciences of south west failure, if it's easier, maybe the listeners will then also be able to resemble that better. Actually, it's a cool story because we did not study at the same faculty of this university. So also in different towns, we didn't know each other before our process mining journey started. I am from Hagen also where ENERVIE groups is and I was studying there economics and informatics. And Jannis is from actually from Mondsee, pretty much I don't know if any listeners will know that place, but it's really beautiful. I think he will also say it's like that. And he was studying there business administration with informatics and our professors just asked us if we want to do a project because you need to do an internship at the end of your degree. And so that way we both came to the ENERVIE group.
03:09
Jakub:
So this project that you did, was it I'm really just wondering here how you entered into the field of process mining. Was this project somehow related to business process management or process mining specifically, or was it just a huge coincidence that you ended up in this field.
03:28
Jannis Nacke:
Actually, it can be said that it was a coincidence because when we were looking for was just a practical pilot project and we brought the broad knowledge of how to conduct such projects. Also Agile projects. Furthermore, we were into informatics, but we were no process mining experts at that time, so we had to learn everything from scratch.
03:56
Patrick
So did you guys have some sort of idea what process mining was before you started this project, or did you just have to do a crash course somewhere online?
04:07
Max Aberman:
Well, in my case, actually, I was working as an assistant student at my university, and my boss at that time, who was a scientific employee, told me about this project and said that doing process mining is really cool. If you want to just look at Celonis maybe before and then you'll be ready to go and then, OK, I thought, OK, seems cool. Seems like a cool topic. And I actually went on to do Celonis data engineer course on their websites because they have those free lessons. And in my case, it was just really important for me. I saw some YouTube videos and I thought, how does this data get into this form? Because it was more like a marketing video, what I saw on YouTube. And for me as a technician, I just wanted to know how do they get this data in this form? Where does it come from? Is it automatically extracted? What do I do? What will I do to have have to do in the project actually? That was a really important question for me.
05:09
Jannis Nacke:
For me it was the opposite. I had no clue about Celonis, no clue about process mining, because my default idea was that I will only need the project management. But then I little bit it got more into the technical details while being in the project.
05:28
Jakub:
So listening to these paths of yours, we recently did an episode about how to become a process mining architect or how to even get to the field of process mining. It was an interesting episode. There's a lot of good feedback on it. But now would you be able to maybe give advice for, let's say, people who are just at the beginning of their career on a how to get to the field based on what you've been through, maybe some general recommendation?
05:55
Jannis Nacke:
I think what is really important is the general interest of obviously in informatics, but also in business in general, because in probably every business lecture at some point you will have business processes and in the end companies just based on different kinds of processes and then with the technical understanding, the IT understanding, you are able to really grasp it really fast and understand how process mining in general works But another suggestion that I have to make is that Celonis is providing this online courses and to get a really broad and general understanding how it works not into detail but in general how it works. These courses are really awesome.
06:41
Max Aberman:
And I would also add, if you're not, if you don't have a degree in informatics or have learned informatics and programing before, it's also not that big of a deal, actually, only because Jannis and me have done it. It doesn't mean you have to have had this because we also are working with was a girl who studied engineering and economics and only had like one module with informatics in her degree and she's also working with us doing automation with Celonis process mining since all of our one year and really likes it and is also good at what she's doing. So I think especially when using Celonis process mining it's really easy to get in if you're interested in it if you have fun analyzing processes, doing automation and everything else and if you just put the work in to the online courses and maybe have a real idea of a project, maybe some data from the Internet, if you're not in the company or have internship and just have fun with it, just play around and at some point you will you will be a good process miner.
07:51
Jakub:
Yeah, I think this is really good lesson here that there is more than one path into process mining and you don't necessarily need to be an engineer or an I.T. person to actually correct this hole and get into it. However, moving on. So you currently work for a company called ENERVIE. I guess it's probably a good to also introduce the company. So could you tell us what ENERVIE is actually doing?
08:19
Jannis Nacke:
Yeah, ENERVIE is an energy supplier in the west of Germany. It is a mid-sized company. We have a little bit more than 9000 employees. So maybe for the Celonis scale, it's pretty small and what we are doing at ENERVIE is just the implementation of Celonis and also the further process automation.
08:42
Max Aberman:
And maybe to go a little bit deeper into into the field of energy supply because before Jannis and I came to ENERVIE group, we didn't really know what are they actually doing? And now we know that they produce energy themselves and then put it to the customers, so sales of energy, pretty much they are trading energy on the market when we have too much or too little in that case. We also own the net structure where the gas and water and everything goes to the customer in the end and the meter at homes of the people is also in our control. So the really broad structure, everything that has to do with energy in our region is pretty much handled by us.
09:37
Patrick
So what is your exact current role at ENERVIE like? What do you guys do? And is it from this project, this initial project that you guys had to do that you got there, or how did you end up here?
09:50
Jannis Nacke:
Exactly how you put it. So first opposition right now can be said as process mining lead at ENERVIE grpoup, so we are currently looking at all the processes and we are implementing all the processes. We have more Celonis experts, but they are not, they are just specialized and we are working for every single department. How did we get there? So the pilot project that we have been talking about, it was in the procurement and at the end of a project it was a very new approach to the company. It was a pilot project and also the way we conducted the project in an Agile way was very new to the companies. So therefore, we had to present it to a boss and then we had to present it to all the bosses. And then it went on until the very top of our company. And at that point they decided that it would be great to solve some issues in our power grid or in the nets with using process mining.
10:54
Jakub:
All right.
10:55
Patrick
So essentially you guys came in with a pilot project to process mining was not existent, and it went all the way up the food chain to the top level. And now you guys are spearheading the process mining lead?
11:07
Jannis Nacke:
100%
11:09
Jakub:
It's pretty cool.
11:10
Max Aberman:
One can also say in our first project we had a lot of freedom. We were the agile project managers, we were the operative project managers in that case but we also had the head of procurement who was like the product owner. He defined the product backlog with us and pretty much steered the project in the right direction together with us. But he had the lead pretty much. And after that project, as we came back as full time employees, the freedom really increased because they already knew we can be trusted. And we were the experts on process mining in different departments and since then have shaped our own way, our own vision of doing process mining at the ENERVIE group. Of course, which huge department will come next is a decision that that not we make but the department and all the important figures at ENERVIE group. But we also can help with the decision and in the end do the work together with Processand.
12:16
Jakub:
Now this is very interesting that you were saying and that was that you bought the trust. And what I find sometimes very challenging is actually to get exactly to this point that you buy the trust and that you persuade these people in your company that first of all process mining is interesting and actually can help them, and second, like a get the ownership or get the sponsorship from let's say the management. How did you do that? And did you also experience any pushback from, let's say, more senior people who just looked at this with a little of concern or distrust? And how did you get over this hump?
12:59
Jannis Nacke:
I think, which is what was maybe the key point here was that the employees, especially the operator employees, were benefiting from our first implementation so much that they said when we did our qualitative interviews for a bachelor thesis that they said they could not work anymore without process mining as a long process mining and the way we implemented it. So therefore and this was already crucial and on the other hand, if you're able to solve problems fast in a way that it's also cheap at the same time, this is probably already enough to make the decision.
13:41
Patrick
Yeah. I mean that's the perfect combination, right? You have the business users being on board from right away and you're solving major issues that is worth the investment of setting up the tool, right? So that's the one of the biggest things that you need to do.
13:58
Max Aberman:
To add on to the question of Jakub, one has to take into account that it's not always easy to convince people of the process mining solutions in your company because the typical scenario that you will meet or encounter pretty much, 100% at some point is that you meet people that say, I have SAP, I know everything about my SAP system. I'm an expert in that field. Why do I need this tool? I can already get all the knowledge that you can get from your process, mining solution and implementation that I already at that point in 10 minutes. Why should I implement a new tool? Most of the times they will still help you implement it, but if you don't convince them to have good change management, you won't get them to use the tool actually. It's rather seldom that you have this problem. Most of the time, if you have a good process mining implementation, the people will be really, really glad to have the tool because you have to, you know that not everybody in the company is an SAP expert. Most of people will really profit from an easy and fast way to analyze that data.
15:19
Patrick
Yeah. That was going to be my question because it seems that there was a lot of enthusiasm from the get go and I know I can speak for Jakub that we have built a lot of solutions over the years and the response can be tepid, right? So very lackluster and people just don't care. So how do you how do you really engage your users to use the tool? How do you generate enthusiasm for what you're doing? Is that something that the users automatically just found out? Or did you have to do something extra? You mentioned good change management. Can you talk about that?
15:52
Jannis Nacke:
For sure. So while implementing the processes we already figure out a lot of pain that our colleagues have among pains in the process itself. And pains that are in most cases, easy to solve with process mining. So what we're generally doing is that while implementing a new process we are already looking for some solutions that we can deliver, but that our colleagues don't know that we will deliver them. So at one point we will invite them into the Celonis for the first time. We will teach them how to use it, which is very important for us, is that we do it one on one. So usually we have them steering it and in the second point then we present our solution to it and they just got a solution for an issue they had for years without even asking for the solution. And at that point the enthusiasm starts. And another point of the change management. What we're also doing is that we are providing this basis to have discussions on the process without finger pointing itself, which is also very beneficial.
17:07
Jakub:
Very well said. This pointing to someone. I also notice that sometimes there is some pushback from the people responsible for the process because you're confronting them with the reality of how poor the process is actually running. And I was just having this example today in my head. If somebody was measuring how I work on my computer, how much time I spend in my emails, how much time I spent on Facebook each day and then told me, OK, so here is like a 40% potential for improvement. I would be like, let me live. Just leave me alone. While if you actually don't point this out, you know about it, but you say, OK, if you do this, you will improve that part. And that's crucial and I really like this approach. Moving on actually. You said that you are currently like spearheading the initiative of process mining and what we usually call it in process mining communities that you create the center of excellence. Is this also how you run things in ENERVIE and maybe if you could tell us a bit more in general way how you are running the show currently.
18:17
Max Aberman:
In our case actually it's pretty much a center of excellence in the ENERVIE group, it's called the Center of Competence where Jannis and me are placed. It's not a real division, it's rather imaginary, but the people actually meet up and talk about the digitalization of different processes in the company. And in that case, only Jannis and me and our new colleagues who are doing automation for us are from the Celonis Department and pretty much Jannis and we are having an agile approach where we work on different projects at the same time, meet up on Mondays to have a check in and think about the most important topics for the week and to have recurring events like true fixes with every department every week or every two weeks, where we discuss with the domain experts what our current developments, how are our KPIs going? Are we getting better? Are we getting worse? Do we have new important topics which are coming from our domain experts which have to be handled? Maybe new analysis is needed, new KPI is needed for the management and everything else, and we discuss that. And based on that, we prioritize all the tasks we have and work on it in a two guys tag team matter. I think actually now four people with automation and what is really important that is that we can also at some time get help from Processand to manage all the projects that we have at the same time because it's not that easy if you have so much departments with not much manpower.
20:01
Patrick
So you mentioned in the center of competence that it's a digitalization of processes. So I'm assuming that the process mining aspect is only a part of it. So what other things are you looking at? Are these projects or initiatives more mature? Was it also spearheaded alongside process mining or what what's the set up there?
20:22
Jannis Nacke:
So actually the Competence Center for Digitalization started with three pilot projects. One was it was I think in 2018 one was just digitalization of documents using nice documentation, document management software, on the other hand we had robotic process automation and then the last topic we had was process mining, which was also the pilot project that we did in our bachelor thesis. And this was the early stage of this center of excellence. And right now all the pilot projects were implemented or decided to cancel them in the future. And currently we are evolving into new projects or we have already involved into new projects, but so far there's only been like a real constant and that was Celonis Process Mining.
21:19
Max Aberman:
Yeah, the other projects are more timely, timely based. So like a generic workflow solution from several departments or a digital way to sign documents. So rather smaller projects which are also really important but not like a huge Celonis execution management system in our case.
21:39
Jakub:
You guys also mentioned that you are working with implementation partners and it's Processand, so that's a that's a very good ad. However, it's not intended. If you could tell us a little or for other companies and other business users that might be listening to the podcast what advice would you generally give into outsourcing for some of the stuff that you need to develop? Are you also trying to balance things and also like get your own experts on process mining so that you manage most things in-house? Or how do you go about that?
22:17
Jannis Nacke:
I think it can be said that we always have the same approach. So in the beginning we do the implementation nearly 100% with our implementation partner in this case Processand. And then after the implementation, the first implementation is done, the validation is pretty much the part where we step. We all of course we also help the implementation partners within the process of implementation in order to deliver the answers very fast to the questions they have in order to have a proper implementation. But we we pretty much step into the process in the validation point. And then there we have a very close, close way of communicating with our implemintator in order to solve issues very fast and efficient. And we also try to cut like the iterations until we get feedback.
23:18
Max Aberman:
I think in the case of the ENERVIE group, the critical success factor or in all cases I would assume is that you have in-house competencies for digital solutions. So it's really important that people in your company actually know how the software works, can make subtle changes, little changes, or maybe bigger changes because in the end, the management won't want that. You have suppliers working for you all the time. You don't want the procurement to get you new purchase orders created every week for more manpower by suppliers. You want a a good balance between work and your own company know how in your own company that way it's not lost and helping with our in-house know how is also so we are in the interface so we can help the domain experts to not have spent as much not have to spend as much time on the process mining implementation because time is the resource that they are lacking on the most if they need process mining to improve their processes actually. So we help these guys and we also have help Processand by delivering the information in a form that is technically easier to implement than domain experts who don't have as much competences in the data science field.
24:42
Patrick
This is a very, very interesting point. My question was going to be how awesome are we as a company, no, I'm joking. But I like and my question is going to be, if, you know, you have experience working with implementation partners, what advice would you give to other companies doing the same thing? How does best set up that cooperation? Right. Because there's some requirements that need to be met and there's internal things that need to happen at the company so that the implementation partner gets the most out of the time spent with the company. So what advice would you give how to best prepare a company to work with an implementation partner?
25:17
Jannis Nacke:
So when first starting, of course, you need to have the key employees at the right position so for sure, you need someone that is able to deliver the IT information regarding, for example, the SAP or how SAP is running in the company. But on the other hand, you also need like domain experts that understand technically what we want to achieve through process mining. And then you need someone in our case it is Max and me that is able to kind of translate efficiently the requests of the domain experts. What is also important is that for example, center of excellence can be seen as a constant in the corporation, a constant that is not switching. So it's not possible that, for example, within a project the employees of the Center of Excellence change if the employees of the implementation partner change, that is not big of an issue. But you have to have someone you can always rely on. You can always ask the questions to.
26:33
Max Aberman:
And to add to that point, I think the most important point is communication because it's really important that the implementation partner and the company are constantly talking to each other. How far are you with the implementation? What problems are there? We have new requirements. Just on the fast route. I think that's important because otherwise the implementer could build something and then it's finished after one month and nobody talk to each other. And it's not what the domain experts actually wanted or domain experts did not work with Celonis enough to actually understand what can also be done and how could it get done otherwise and that's a set up for not not not for failure but for not such a good uh, project at the end.
27:27
Jakub:
Yeah. Talking is the key as with every other relationship. Guys, moving on, let's get a little specific. So could you tell us what kind of processes so far you've implemented? And since we know a little, where are we going to head, we will then focus on one specific.
27:46
Max Aberman:
Yes, it's already really many processes actually. So we started with purchase to pay and accounts payable. I'll just list them first and then we can go into detail in some of them, um, after that we started to look at the meter to cash process for our net provider. I think it's called in English. And after that we started to look at our sales department and in that case the analysis of open requirements. Uh, yeah. So we sent an invoice to our customer and he has to pay it. So if they don't pay it, it's open. So monitoring our money that we need to get from a customers actually. Accounte Receivable, thank you very much. And we also started to look at our controlling process. So budget planning for the departments which is actually not as easy in SAP if you're not working often with it. And I think we will also have a lot of success with that one and I don't know that our really many processes in which one do we really want to go into how much detail?
29:08
Patrick
Well, I mean I was going to ask I mean, it sounds like your process mining is fairly extensive. You're looking at a lot of processes and what was the rollout like and what was the time it took to implement each process? Was it all at once or how did you structure it?
29:22
Jannis Nacke:
So like we told before, the first process was the pilot process, uh, of the procurement and at the same time we also didn't the accounts payable, this was just because the head of procurement was very innovative and he said he wants to digitalize his job. And then from that point on we started to go there where the issues are the biggest. So especially in the grid, we can say that it is pretty hard to always keep up to date because there are certain regulatory approaches that you have to fit and there are new regulations that you have to follow every single year. And at the same time, we also have had an SAP migration and every listener that had an SAP migration might know that this causes some issues, sometimes, just some...... So this is where we went and afterwards we continued with our accounts receivable. Probably because it's also one of our cash cows of the company. Yeah.
30:40
Max Aberman:
And to add to that. How much time did we need for those implementations? So we started with procurement and accounting and did the basic implementation in five months. After that, we wrote our bachelor thesis for three months. So we weren't working at the ENERVIE group, but we also assisted at steering in our next projects. Which is the best to start after the start of our actual employment. And that was a net. And we are working with the network or grid provider since April last year. So more than a year now and there's no end in sight. We have much use cases in our backlog, much of important topics. So it will go on for forever maybe. I don't know yet. And with our accounts receivable process we started actually last year, August around that time. And that case we have to say is it was much more complex than we thought and we had some really big problems implementing it too. So not every process is easy to implement because SAP logic is sometimes strange or and sometimes totally not understandable. But we did it actually. And we can now say if a position in our system was actually cleared or open. So that was the most important thing for the use case. And we actually implemented that. So like six months from now or seven actually. But the project is still running and we also have some more use cases in the pipeline and for controlling in that case, Jannis and me did not assist much in that case because Processand did everything like in a month and then it was pretty much ready. And at the moment we are validating with the domain experts. So that one was really, really fast.
32:39
Jannis Nacke:
Yeah. One thing that we need to mention is that the processes are running that long because the first implementation is very fast, usually within a few months and then the process of change management starts and the change management is way harder than the actual technical implementation. Yeah, except for the last project that Max just mentioned, there was an implementation also very hard because we had two source systems two SAP systems that we have been technically combining that would be using different logics and to get both of them on the same side was pretty tough.
33:21
Max Aberman:
And to add to that, we were also looking at positions and invoices starting from 1999 in one system and starting from 2004 in the other system. So yeah, things changed over the time and it's not that easy to look at 30 million positions or more or actually it's 70 million at that moment in our Celonis if you click on the dashboard.
33:46
Patrick
Yeah, that's, that's quite a big data scope you have there. And that kind of leads me to my next question. I mean with all these different processes that you're implementing, all these different views and KPIs that you need to calculate, is the data pipeline an issue for you guys? Was this ever a concern? You kind of mentioned of adding to source systems and that kind of caused some issues. How is that handled?
34:09
Max Aberman:
So if you mean data pipeline, you mean problems, with the extraction and everything getting all that data into the system. So yeah, actually it happens that some data lots fail or some extraction will fail. That that is a huge point why we are currently thinking about going from extraction one time a night to realtime extract to make it on the one hand more usable for operative employees who are looking at cases and then working on them live time. And they also want to see that the case is finished like some minutes afterwards and not on the next day. And another point is that it will make it way more secure if we just load smaller chunks every other hour or every other minute, than loading everything overnight. We have actually had the case that two times it failed and like the last month or two the data load failed in our meter to cash process. And one operative employee just right away contacted Jannis and said, what is going on? I don't see my problematic cases in my inbox? How do I work now? I don't want to go back to the old times. So it's really important for us that the data pipeline is in check and everything works as it should. So that's a huge point when doing process mining to set that up correctly and have it in a secure way.
35:43
Jakub:
So one of the processes that really caught my attention was meter to cash, which I think is very, I would say, unique or specific for energy industry. Could you explain to our listeners what meter to cash is and what are some of the specific problems and maybe use cases you're trying to solve there?
36:06
Max Aberman:
So the main point in our meter to cash process project was process transparency. And what is the process really about. So as it is, it says, meter to cash, we have a meter which is placed at the home of one of our customers or at an industry plant of one of our customers. In that case, there are different kinds of meters. So you can have a meter which is, which is counting the energy that you consume or the gas that you consume. But we also have customers who have solar panels on their roofs and produce extra energy and put it into our net so at sometime we get money from them. At some time we actually pay them money for their for the energy supply. But in that case, we don't do that all by ourselves. But we have to communicate with the energy supplier because in the meter cash process, we only looking at our grid and therefore we only get money from the use of the grid from our suppliers. So it's different market roles in this case and if we come back to the basic process, we have a meter at home or at the plant. We need to read the the values on that meter. So how much energy was consumed or produced? Get that into our SAP system. It is stored in a meter reading document and after that we go into the process of market communication. So we need to provide the energy values, the measures that we got from the meter to our supplier. He has to answer to that message in a special kind of way that those are the regulations that Jannis was talking about, so the communication is really just you have to send a specific segment in a specific form and if it's technically wrong, you get an error message and everything else. There are lots of messages sent to different suppliers every day from us and many responses we get back in that case. And if all those processes with those messages go right, we can actually produce what is called Netznuchenabrechungen in German. It's the invoice that we create for the usage of the net that we sent to our supplier because he has used the net. And after that the invoice just gets cleared and we get the cash in that case. So from meter to cash.
38:35
Jakub:
Before you go into use cases, I have one very specific question before, because before I worked with business data, I worked with major data because in my past career, I worked in Siemens in building technology division. And I know that the biggest problem was always to get the data from the meters at the customer side to your data because you know, you have a meter somewhere in the area where there isn't even Internet and you are relying on like Bluetooth and sending data somewhere through the air. Is the data consistency an issue for you guys?
39:12
Max Aberman:
So we have to differentiate in that case because the process that we are looking at, we don't have many smart meters who actually sent the data automatically. Until now, the process of changing to smart meters in a lot of cases is actually going on right now. And we are also assisting with Celonis because we have some data problems on specific cases and we have to have the perfect data setup before we can change to smart meters. But that's a different, different story in the book and that case because normally if you have a meter at home, our system actually now also Celonis what was beforehand done by an employee of the department, Celonis classifies whether the meter is going to be read by an employee of us which goes there and just looks at the value or if we sent a card by post?
40:13
Jakub:
It is post.
40:14
Max Aberman:
Per mail, ok, sorry for that. Per mail and the person reads the meter by themselves and sends it back to us per mile, per app or per mile again and in that case, we get the data like how it is seen on the meter. But you can understand that not everybody will answer our mail and not everybody will let us into their house. And that's a huge problem because if we cannot measure the value on the meter, we have to approximate it. So the SAP system makes an approximation, but there are many ways and many problems which can block an approximation though in that case, a meter reading document can stay not filled, which is problematic because if we don't have any values, we can't actually place an invoice and can get the cash in our meter to cash process and a further problem with that.
41:14
Jannis Nacke:
Yes, it is creating a lot of rework. So if we approximate the values from the customer, then there is a possibility that the customer is coming back at us and saying, excuse me, I had like the 0.001 kilowatts less consumed and then the entire system has to be changed. The entire data in the SAP system has to be changed and the process is starting from scratch and obviously we want to lower the rates of approximations, even though it can help in scenarios where we are not able to read the meter.
41:58
Patrick
So as someone who has ignored many mails and has not been home to let people in to read the meters, I apologize I didn't know this was such a big problem. I will be sure to do better in the future. But my question was going to be, if you look at this meter to cash process, what are the immediate problems that you're trying to solve with process mining?
42:21
Jannis Nacke:
I think what is very important here is market communication. So everything is regulated by the German government and also by the EU, and therefore the communication has to be on point so an immediate issue that we have all the time or that every single energy provider has all the time is that the communication is for some reason going wrong. That can be on one hand because the meter reading didn't work. That can be because metadata change and also because physical meters don't work anymore. So this is probably the biggest issue to communicate in an appropriate way based on a good database.
43:07
Max Aberman:
And you can imagine if we are sending several thousands of messages out and getting several thousands of answers every day, we can't have as much employees who look at all those messages sent or received. So we have to have a really good SAP system a really good automation in the SAP system to not only send the message in the correct format and receive and read the messages in the correct format, but to also handle some of the problems that can happen if if problem is actually there in the process. So we built up a monitor which analyzes those different categories of messages and a possibility to prioritize the cases that are problematic because we also as we started working on that on the topic, we saw with our net department we had cases and problematic cases in communication which have gone back to 2016 and 17 so they're really really old and if we couldn't place an invoice in the year 16 or 17, we also cannot place an invoice in 18, 19 ,20, 21, 22, because that's a root problem some years ago. And you can imagine how important a prioritization in that case is for the employees that are not much, not many actually to work on all those problems which have stacked up over the years. But to spoil it a bit, we are on a really good way and the problems have minimized a lot in that case.
44:51
Jakub:
So you're saying minimized then my question would be did you or you then were able to actually solve some of these issues where you also able to quantify it, let's say like how much you improved certain parts and what those parts were?
45:08
Jannis Nacke:
Of course we are. But I think this is too critical information, but what can be said, first of all, for all the listeners to understand how many customers are on our grid we have around 400,000 customers on our grid every single day. Being supplied by our grid. And this will also increase a lot in the coming year.
45:35
Max Aberman:
In that case, I would also add information type that is not really critical. I think if we if we don't work with absolute values, but with relative values and one can say that we have more than halved or actually in the last month we have only 1/10 of the problems with the communication and the number of missing invoices which have needed to be created was also more than halved. And in that case, you can also imagine every month we have to create new invoices. So it's not a number that we say we are missing X number of invoices. Now let's work on those, but because every day there are new invoices which have to be created and we have to work against time, pretty much like working against the wave of new, new problems arising. And we are having a really good positive trend and then that makes us really happy and our domain experts also.
46:43
Jakub:
Wow, sounds awesome. And I just want to say congratulations because you know, when you start getting these results, it's very easy to buy yourself more into process mining. And this is exactly why process mining exists. Anyhow, I also talked to a colleague of mine who was supporting you on some of the implementations, and she told me that you actually were building this this thing that you called ENERZONA. Can you tell us or what's up with that?
47:12
Jannis Nacke:
Of course, for sure. So when we were in our pilot project in the procurement we have figured out, so we did this classical process mining where in the process the issues created. And in the end, a very logical answer of this was that in the beginning of the process where all the employees of the ENERVIE group of order something a lot of mistakes were made or generally are made. And this is not because our colleagues don't know how OK, sometimes they don't know how to use it, but it's also because the SAP is very complicated and there are many sources where you can make issues and therefore, we created a vision in the end of a project that we would like to take our colleagues out of the beginning of process. So we would like to create a platform that is like Amazon. You just give a name or something, you need something, you need Natrium, you just put Natrium and then you get like the possible suppliers that you can choose. And when you put on it, you don't have to enter all your data every single time. But the system already knows who you are because you have audited two years ago, and this was the vision that we have created, and based on that we got two new employees that wrote their bachelor thesis on this topic. And what they did is they would take the existing process and take the historic data that we have within our cloud and then build a business view in Celonis that would allow to just search for something that was bought in the past. And then and directly just by it's basically two clicks so you put in the name and you click on the object that you want to have, and then you click on audit and then it would already directly order something within our live SAP system.
49:15
Jakub:
So that means that you are basically executing a certain transaction through Celonis implementation in SAP.
49:24
Max Aberman:
Exactly. If we go into detail, it's by using RC and calls, which are used to create a purchase order in that case and all the irrelevant data has gone from Celonis and written into correct fields and thereby a purchase order in that case, for example, has been created.
49:41
Jakub:
Sounds very cool. All right, guys, what's in the future for you when it comes to process mining and all of these initiatives that you're running?
49:52
Max Aberman:
Actually a lot. It would be nice if there were more of us actually in that case because we can't work on all the stuff that we want to. But we, but we have a huge plan actually. The great picture consists of automating more. We have a lot of repetitive tasks in ENERVIE group which are hindering our colleagues from doing the actual valuable tasks in their job. And then they need that time and we are assisting with Celonis, but we only build some automation. So we are really sure that there's a lot of potential left at the company to automate. A further point is this establishing this culture of continuous improvement even more because we have our departments looking at their KPIs at their performance every week they talk to each other, how it developed, how happy they are. It's actually helping what they're doing at that moment. So this culture is already established in lots of lots of parts of the group of the ENERVIE group, but we will assist to bring it even further and a different point is also that we want to yeah, implement more processes actually because we've only looked at some major processes, there are some other left that we can address in the future. So that's a huge point. Another point, so it will be a point that I'm talking about because we just presented this great picture on Tuesday actually to the high figures of the ENERVIE group, to the board. Point number four is actually the establishment, which is the idea of us or our boss pretty much the establishment of a cash flow liquidity management system, working capital management system. We are looking at different data from different departments in our company and just look how our liquidity is going in the different points where we can access that data. So how much a open and not cleared positions we have.
52:22
Jannis Nacke:
It can be said that we are looking to create a cash flow management that is real time. So in every single second the company knows exactly how much money is existing. And I think this sounds not that crazy, but it's actually a huge step because then we can steer more precisely how we spend our money. And since we are trading a lot of energy, which is a lot of money that has to be used and a lot of liquidity that is needed, this can be very, very valuable for us. Yeah. Sorry I interrupted you.
53:03
Max Aberman:
No problem. Thank you for the clarification that point. So it's pretty much the voyage to Celonis EMS. So execution management system. We take data from our sales department, we take data from our net department, we take data from our accounting and combine it in the management dashboard for the people who are actually working on liquidity management and have it all in one place, not in several SAP systems, not in ten reports who have to be executed the whole nights, but just access it and look at the numbers, which are correctly fine. And that will be a huge, huge deal, if we actually do it at some point. It's a really ambitious task, but we are on a really good way in that case. So maybe Jannis, you want to follow up on the other points of the great picture, in that case?
53:55
Jannis Nacke:
I think another thing that we haven't been talking about is that we want to steer our sustainability in the company by using process mining. We have already created the transparency within the procurement department by combining all procurement data with sustainability data. So we know exactly how much CO2 we are currently emitting by buying products. And here's the great picture that we don't want, only want to know how much we are producing, but we want to steer the behavior of the operative procurers in order to decrease our CO2 emissions in the supply chain and before we buy the product.
54:42
Max Aberman:
Of course, at that point we are only using estimations at that point which are pretty good, but are not, uh, not ready for the actual steering of the procurement department. So we have to enrich that data even more. But I think that's also a good point for the usage of Celonis EMS in that case, because we have gotten a different system where the CO2 was shown to us on a dashboard and our head of procurement, which is sadly not with us anymore, but the new head of procurement is also really a really cool guy.
55:15
Jannis Nacke:
Just to precisely formulated he did not die, but he changed.
55:20
Max Aberman:
OK, I'm sorry, I should not use it in English anymore in that case, haha, ok, I will remember. So he's not working with us anymore at the ENERVIE group. But he had this idea, why do we need two systems? We have Celonis, Celonis is great. Our procurement employees need to have that CO2 data also and Celonis and Processand helped us to implement these kind of data also in our process.
55:47
Jakub:
That sounds very cool, guys. I mean, I am a little jealous of having that much of a freedom to work on this because, you know, we are still consultants we are usually limited by the budget. We can give ideas, we can work on different stuff, but we are always somehow restricted and just listening to you and hearing this, this ambition and excitement makes me feel very, very happy for you.
56:13
Patrick
Yeah, same.
56:14
Jakub:
So I guess we will have to invite you back in a couple of years to hear how that went.
56:20
Jannis Nacke:
Yeah, for sure. Thank you so much.
56:21
Max Aberman:
Would be really nice.
56:23
Jakub:
Anyhow, where can people find you if they had at some point, some question or would like to also hear from you? Where can they get in touch?
56:33
Jannis Nacke:
I think the best point is LinkedIn. Yes, well, you can just send us messages. Obviously you can also send us mails. I don't know if you can link them below, but that would be the best approach.
56:49
Max Aberman:
Maybe if you can't, Jannis Nacke and Max Aberman, only one n, as that's an important factor actually, which creates a lot of confusion.
57:00
Jakub:
We will try to put it into the episode description. Anyhow, guys, it's been a pleasure talking to you. I'm very happy that our colleagues who actually worked on the project told us about you and got us connected because this has been a real pleasure. So thank you for coming.
57:19
Max Aberman:
Thank you very much for having us here.
57:21
Jannis Nacke:
Yeah, thank you.
57:22
Jakub:
To you, dear listeners, thank you for listening. As always. You can also write us an email either on LinkedIn where we are very active. So just drop us a message on Mining Your Business profile or you can just write us an email on miningyourbusinesspodcast@gmail.com If you like us, rate us, you can rate us on Apple Podcasts, Spotify basically anywhere where you usually listen to your podcasts and your feedback is always very, very valuable. So, thank you very much and we will be looking forward to talking to you in the next episode of Mining Your Business podcast, guys, thank you very much. Bye bye.
read more
Discover more episodes
GET IN TOUCH
Begin Your Process Mining Implementation
With over 450 process implementations, we know exactly what the crucial parts of a successful Process Mining initiative are
Daniel Bitsch
Senior Data Scientist