Session Information:

Learn why Process Mining itself provides little value

Those who do not know the problem cannot solve it and every effective therapy requires a proper diagnosis.  If you want to maximize the efficiency and resilience of your business, there is no way around digitalization and automation. 

UiPath is not only the market leader in Robotic Process Automation (RPA), but is also the only RPA provider to offer its own Process Mining solution.  These are just two of the components of UiPath’s unique Hyperautomation platform.  In this session, Rudy Kuhn will show how the interplay of Process Mining, RPA and the other components of the UiPath platform enables the sustainable improvement of process and maximization of efficiency. 

Session Transcript:

To our next presentation here, we're having Rudy can join us. ... is the VP of Data Analytics and had process, my name bass that are for UI path. Rudi, please do join us, are excited about having you here.

Rudy was probably one of the first to recognize the potential off process mining after 10 years in business consulting at IBM and other companies. He founded Process Gold in Frankfurt, Germany, in January of 2010.

It's a, it was the first consulting company specializing process mining in 20 16 process called Publish its own Solution, which is, which is used by E Y, Worldwide, as a standard for process audits. At the end of 19 process, gold was acquired by UI path. Rudy is the vice-president of data analytics and process mining, supporting over 7000 customers in the integration of process mining and robotics process automation. Rory, we're very excited to have you here with us today. Very much looking forward to your presentation.

Hey, guys. Thanks. Thanks Jose for, for the nice introduction, guys.

Welcome to my presentation, and I can tell you, nobody is more than more happy than I Me that I can really join this meeting today. I will tell you a little bit later why this is the case. I almost didn't make it, but I'm here and that's, that's great. That's great news actually. Good. Yeah. So introduction enough, you know, of course, you should go by Rudy.

I'm probably read the guy who first recognized what process mining is capable of doing and maybe I was one of the first to also realize what process money can not do. But before we come to this, you know, this presentation will not celebrate process mining is the greatest technology ever, but I will be more skeptical. And I will talk about some of the mutations of process mining. And what is good for, and where we need a little bit more than just process money. But before we start that, you know, let me just introduce our company to you know, one of the very boring slides.

So UI path, probably is one of the fastest growing companies originally founded in Bucharest in Romania, now headquartered in New York. And I'm based in Frankfurt in Germany. We have more than 7000 customers. And what do we do for these customers? Why we do RPA. So, RPA, robotic process automation, I would love to talk about robotics, process automation, illiterate in world. But this, this presentation really is about process mining. So let's stick to the topic. But, you know, we have more than, probably by now more than 3000 customers worldwide. More than one billion of funding. And just recently, we had some great news that we're, you know, that which were picked up by a nest egg.

And you can actually see this in New York on, what is it called, Time Square. Right?

So right now, UI path is worth more than 10 billion.

So we're, we're approached Descartes Corner lot and also, UI path was recognized as one of the most disrupting companies in the world and we ended up on the CNBC list.

I think of disrupter 50, OK, guys, enough about the company: What is the topic today?

You know, we want to talk about process mining and why it's important, why it's important because right now, everyone is rethinking that processes and organizations for business agility, Um, Covert 19 S, a very, very bad thing.

But, it helped really, people and organizations to understand how important digitization and, And, and automation really is.

So, mean, more than ever before. Companies are really rethinking, you know, how can we increase resilience? How can we increase efficiency? How can we reduce the risk in our organizations? Then you normally people working on the place, processes, end to end. So this is really what keeps people, you know, awake at night.

And Gardner recently came up with A with a new term called hyper automation and what is hyper automation? You know, in hyper automation you receive the word automation. It's actually the majority of the letters. So hyper automation is about automation. That's why RPA release and the center of everything. But, if you really look at it, and you think about, what do we need to do become, really an automated companies thinking automation? First, robots, first, really implementing this technology.

We need more than just the pure RP A platform, So our company is the leading RPA provider in the world. But we think it's not enough, and coq not recognize the same topic.

Know, if you really want to be successful with RPA, you need to add a couple of things. You need to add a ..., you need to add machine learning. You need to support a long running workflows. You need to have advanced analytics, native integrations, with systems like sap, like anything else out out there.

And, of course, you need process mining.

First. Before you automate, you should always think about what the process really is. You need to recognize the processes as you need.

First, to understand what the problem is, before you can really tackle it before you can automate it before you can apply AI, long running workflows. And, of course, we need some analytics. And, of course, what you use, process money, again, for advanced analytics about the process. So, when all this really comes together, this is what we, and what Gartner really collects hype automation, And I'm very, very proud to be part of this company, because UI Path probably is the only company in the world that really has a complete hyper automation platform, as we call it.

So, you know, before we as process code joint UI path, there were basically only three boxes, build, manage, and a lot of you know, the blue orange and the red box. So, we had the best platform to build bots to manage faults in the cloud or on premise and to run, or attended unattended bots or test pilots, and all the other stuff we have.

But again, you know, the biggest challenge for customers really wants to understand, OK, what kind of process do we really have? What should we automate? And after we've implemented automation RPA, how is the process now improving? What's the impact of RPA on the process? And how can we engage with the people?

So, this is Why ... process code, We also acquired a second company called Snapshot. So, now we have a couple of discovery products called Automation Hub.

Automation. How basically is like a social network where everybody who was working in their organization can propose automation ideas, this will be picked up by the Center of Excellence and would it be implemented.

We have test capturer.

This is like, a, like a bot, or, like I said, like an ancient watching over your shoulder digitally and recording what people are doing on the front end.

We have process mining for the end to end backend processors, you know, the business data, and we have task mining, which is currently in the private preview phase, where we can better understand what's happening in the front end. You know, the front end is really the place that was kind of blind spot.

Or process money in the past, because we never saw any digital footprints from applications like e-mail, like Exxon, you like, like, like, PowerPoint or, you know, that that's well demonstrated a charge.

So now, with the combination of the different products, we can really better understand what's really going on front end, backend, we automate, We measure, we monitor, and also, during my presentation, I would talk about our latest integration of Process Modeling and Action Center, Because this really is where everything comes together.

And, know, so, what does process mining? You've seen couple of many presentations, or you about process money from me? The easiest way to describe some, You know, to, to explain what process mining is. It's like an x-ray system for business processes. So, if it hurts, or maybe if you want to avoid paint in the future, you take the data, you approach the platform, you push the magic button, and you will get 100% transparency. And if you compare this approach to the, to the tradition of the work I did for many years as a consultant for IBM, you know that the workshops, where you end up with more processes, when you have people in the room. Because everybody's telling a different story.

You will save a lot of time and you will save a lot of money And no x-ray is some very, it's a very interesting topic and today I want to share let us story of a few, which just recently happened to me, unfortunately actually last week. So what happened last week?

This guys, I love biking so I go on a bike ride every morning for something like an hour. Angela, last week on Wednesday, very bad luck. I hit some unexpected speed bumps in full speed.

I was catapulted over the power of my bike, and I actually ended up lying on the road. I was lucky. I have wear helmets. But, yeah, I actually fell on my arm, my shoulder, and my elbow. So, I was delivered to the hospital, or brought to the hospital taken to the hospital by, by the ambulance, and they X rate my elbow and my shoulder and they realized, Oh, it's broken. It's not only broken. It's actually a little bit shadow, but there wasn't really, you know, very, very nasty, just last week.

Now imagine that, the company would stop there and just telling me, Hey, Rudi, that's the problem. Here are the pictures, you know, take it, take it home goods like all the best and have fun.

What that solve the problem?

No, of course, not, you know x-ray diagnosis is so important to understand where the pain and where the problem really comes from and I can tell you a show that shouldn't like shouldn't look like this, but, you know, what they did. They actually fixed it.

I know it looks messy, but it leaves me. It hurts, but now, it's fixed.

That was, yeah, three days or four days.

So, I'm more than happy that I can join this session today, because I wasn't sure if I can make it, but now everything is fixed. I don't have a cast, I have to move around. But that's the difference, guys, between diagnosis and solution.

Of course, we need to understand where the problem comes from.

But after we understand the problem, we need to apply a treatment, and you know, of course, being a business consultant. I know there are many different ways our problem can be solved. You know, sometimes just a little bit of training might be the right way. But RPA is a great way to increase efficiency and to reduce risk, OK. Let's move forward. And let's start, you know, maybe if the first part because I would like to show you really in a practical way how we are using our technology, our process modeling platform, to discover RP potential. So, now we're going to apply the x-ray and see, you know, what the problem really is.

For that, I switched to my browser.

And what you should see now is our process mining applications. So, you know, since 2010, I think we have to live on something like 500, 600 process mining projects. And over time, we realized that not every project, or processes really different. So, we created a, sort of, a template and application template. You can upload any sort of data into it, and out of the box, or get 50 or 60 different dashboards with all kinds of insights.

And today, I will show you only 3 or 4 dash plus, because we don't have more time. So this first dashboard is perfect for the most boring dashboard we have, because it can only give me some indication of the data we have.

And we can see, you know, we have 20,000 cases. And a case in our example based on the activities like received Envoys process invoice, proofing wasn't paying invoices. Apparently this process is about invoice approval.

And the interesting thing is that this company, is it, by the way, it's some real data from one of our customers.

And they use the HTTP workflow system for the invoice approval process, and they really figure out 311 different ways. We call them variants how this process can be done.

But let's go back to the first quarter of 2019, because this is where the automation journey for this company started. So let's take a look at the first quarter of 2019. So what do we have in the first quarter?

The first quarter, we have 6000 invoices, and we still have 165 different ways how this process was wrong. But let's take a look at the process. So, when I click on Process, the system will show me the most common way, how this was done.

Know, we have this box, as we call these Boxes Activities, we have to the connections, the connections, are the edges, And because both of my sliders are set to 0%, already, see, only the most common way. you know, basically, very number one, how this process was executed.

So, if I click on bearing number one, the system will apply a filter, and you will see, we start with receive an Invoice Processing Voice, We finally check it, we approve it, and we pay, But if I do not apply any filters, of course, I can add more, and less common activities.

Now, we see that most of the time, go straightforward, but sometimes, actually, for 1796 cases, we take those sidestepped, we first request some data. We check the contract, and only then we do the final invoice.

The numbers you see on the screen, of course, they represent the number of invoices.

We can easily switch to other metrics, like, for example, the case amount, And you might be noticed that the color changed over here because, you know, if we have a bigger inverse, we had, uh, we have a different approval process. Well, that's the way it is.

Let's go back to the number of cases. And if I add more edges, the system will of course indicate that, for example, permanently define check off invoice is not find all the time because we see some groups. So, this is not really that this doesn't look really efficient.

And if we add one more exception, we might discover that we have some compliance issues where cases are bypassing the approval step, really doesn't really happen, but we can check, you know, we simply click. And if you'll click on it, you apply a filter. For all cases where the furniture was historically followed by paid invoice, we approve our our selection.

Oh, and we see that some cases apparently come down, only to be approved afterwards. Oh, and we have a great franchise. To visualize this, let me show you.

It's the the animation.

So, if I turn on my animation, you will see this little dot over here, and if I start the animation, then every single dot in my presentation on my slide, my movie, basically represents an invoice, or some are moving very, very slowly. Some are even standing still. I can always stop it. I can zoom in and, for example, ask, Hey, what are you? What's wrong with you?

And if I click on, the system will take me to the detail type and will show me exactly the master data with an SLA violation.

So, that's very That's a unique functionality. We have the tagging, I will talk about this in a moment. But, we can exactly see, you know, it took us 83 days to pay this invoice, and we see exactly when, you know, the balts, other people, what they did, how long it took. And, if we have the right car, right data like here for the system. We can see actually, cost us $7, But. let's go back to the process.

You know, if I would really like to know if we have any cases, without any approval, I can also filter for cases where a proof invoice was never, actually, cases never passed through approve invoice.

We create another filter, we select our or, we confirm our selection, and if I add all the complexities of the process, we see the process, and we can look. you know, it happens very often, very number five.

So, this is what varied number of life looks like. Basically, who cast?

I would like to understand water suppliers and maybe what's the value, and how did the value changed over time, and maybe less responsible. So, this component is very flexible. And you can always, you know, configure it in a way you want, without being a developer or editor. So, if you would like to change the sort order, you simply drag and drop this guy behind the case on, or you can. You can go up one level to supplier type, and so on and so on. So, you really get full transparency of the process.

But this is justice. No stand-up process mining. This has little to do with, with RPA.

And, of course, it's great to better understand, you know, if I remove the supplier type, or we can see that this happens very often for environmental services. So, again, one more click. I approve my selection, we're down to a 370 cases, were almost one million, And one more click on this icon. We jumped at the details, and we see a list of the 370 cases.

We can sort them case amount, we can sort of throughput data. And if I click on any of the cases, I would again see exactly the consort diagram, what happened when, what kind of no exceptions we had, and so on.

But, no, this is just pure process model. This will not help you to solve the problem. It will show you where the problem really comes from. But if you want to solve the problem, you need to do it in a different way.

So let me show you what we can do. if I remove all my filters.

I go back to the variance.

Then I remove this one and we look only the number of cases.

What, so what can we see that we see, when I go back, we see 165 different variants. We have 6103 invoices.

And, know, I'm, I'm fairly new to the RPA world. But my colleagues really told me, Rudy, there's no way. We can automate this mess. There are too many exceptions to many variations. That simply doesn't work. If you think about a bot who can really manage this, it will be too complicated to program it.

But what about the five most common variants, or if we go for the five most common variants? Wow, look at this.

I would say this picture looks so much better. And we still have almost 5000 cases of 44,000 HIV 70 cases.

So 70, 80% of all cases are handbooks, or, you know, are covered by the five most common variants. And I'm pretty sure we can apply some automation here. But where should we start?

What what about the bottlenecks?

So when we switch to the average throughput time, the system will indicate, again, based on the dark kind of coding, that we have a huge bottleneck between receiving wasn't processing voice of nine days, because these nice days are sick are significantly longer than any other part of our process. And we have 1.3 days here. We have one point four seconds here. So nine days really seems to be the bottleneck. So let's do something about process invoice. To better understand what's really going on in process invoice, we need maybe we're not where we have in our toolbox called Task Mining.

So, let me introduce you to test combining.

What is test marketing? You know, ..., we basically, we know, again, with process discovery, that's the, the overall process, mining, and audio tools, we have at UI path. We want to understand how a process is really executed, and if you look at the process like sap purchase to pay based on the data from process mining, you can only discovery, based on the big milestones from sap, for example, how the process is run. And you can discover what should be automated.

Like for example, you know, we discover we have a bottleneck, we create purchase order, but before we can automate, we need to better understand what's really happening over here. So, what we want to do, we want to focus on create purchase order only. And we use our task mining capability to understand not only what to automate, but how to automate it.

And task manager will actually tell us that create purchase order in this example means you will receive an e-mail, You create a PDF, or you read a PDF, copies some data, and only after you save the data in sap, you will see the time-stamp for create purchase order or a year.

And that's the level where, you know, automation really takes place.

So, how, how are we looking to combine process mining, text mining in the future?

You have seen a similar picture like this, here. So, we see we have this process in voice over here. And it's a bottleneck.

And now the system is asking us to select a task on the left to visualize its execution. OK, let's do it, so, you know, let's say I want to understand, for example, how process money really works.

So, when I click on it, based on the data from task money, you know, the data we collected in the front end of an application, up front, and of all applications on the desktop.

We can see that process invoiced, basically are two different ways.

This is the most common way. But there's a second way. So when I click on it, we can see, Oh, look at this. So this way of the screenshots, we see, urging what people did.

And we see, oh, this is less common but it looks much more efficient.

If I want to automate this one, again, I can click on export the studio. And the system will export an XML file a separate file directly to the UI path studio where I can take this as a skeleton for my bot and immediately within minutes create my bot to automate process invoice.

Now let's assume with automated it. So what's the impact, I want to see what really changed. And for that, I go, again, to back to process, mining, to a different dashboard called Automation.

And here, again, I add or complexity, we're still in the first quarter of 2019, so you know, this is where, where the automation, or before the automation, Jodi, for our customers start, just let me switch. So what do we have? We have the same process, but the kind of coding represents the automation rate. So we see a legend over here with zero to 100%.

So we can see that already in the first quarter of 2019 process envoys was fully automated by 100%.

Same for pros for pay invoices, or there was an interface to the payment system, 100% automation, But everything else in between was completely manual.

And we have this big bottleneck. And you can see that, for example, every execution of our most common variant, very number one, in the first quarter, for the five most common variants, cost us $98. Well, that's pretty expensive.

So, in the second quarter, this company started to implement UI Path. So, let me show you what happens.

And in the second quarter, first of all, you maybe notice that process invoice has a slightly different color, because now, by the end of the second quarter, it was automated. By 20%. 25% of all, invoices were already processed automatically.

And we already re-used a throughput time, if you remember, from nine days to seven days and look at the cost. So the cost already came down from 98, 197.84 to roundabout 81.

And in the, in the second, in the third quarter, it gets even better. So, if I click to the, if I go to the second quarter, you can now see that what, it looks like processing Western of fully automated by 100%, which is great.

For some reason, we have even started to automate a bunch of invoice by 34%.

Look at the, look at the throughput time, you know, we came down from eight, not for nine days to seven days, amount of 34 minutes only.

And, well, look at the costs. What do you remember, 98, 81 and now, $30 only?

So, great accomplishment. Everybody, you know, mission accomplished. Everybody, happy, not quite. Let me show you why not?

I go back to the first quantum, and lets you gentlemen for, now, please focus only on this part of the process. You see, we have the 1.6 days over here.

And we have the 2.2 days over here, in the first quarter, in the second quarter.

In the second quarter, after we started to automate this part, we increase the throughput time by Harvard data 2.1 days, OK. Nobody really noticed. But people notice that this part actually, you know, from, I don't want to five days for something like that, or three days up to seven days.

So that's the reason why, in the third quarter, the company started to automate a bunch of invoice, whereby able to bring down the throughput time over here, I can for four days. But because of the higher workload, you know, this was much faster mill, there was a high workflow. The resources over here couldn't keep up with the work. So we have a new bottleneck here with four days or four days again, here.

So, I have to admit that maybe this dashboard is, it's not perfect, so we're working on some really cool new solutions and new dashboards where this would really jump into your face, and that way, you will see this much better way.

But, no, I hope you understand where we come from. So, first, we take the data.

We connect directly to JJ. That's something I didn't mention so far. We have complete built-in data transformation layer. My baby, maybe we don't have enough time to talk about this.

We, we, we, we visualize the process. We discover RPA potential. We use task mining and other functionality to really better understand what process invoice for itself, it really means. And after we have implemented UI path RPA, we can measure the impact on the process. We can measure the increased or decreased throughput time. We can measure the decreased costs. And we can really show the success and the impact of RP on the process.

That's cool, isn't it?

But that's not all. Let me show you two more things.

We have something we call. When I go to my conformance dashboard, you know, we don't want to stop with measuring and discovery.

So, when I go to my conformance dashboard, we get, we get a dashboard called techs.

Text is a unique functionality we built.

We built in, in our applications, different to, I think, all of the other applications in the market. You IPF process money.

We cover the entire Data Transformation layer. So maybe let me show you this before we continue here.

We consider this really to be one of the big highlights.

So, you know, like in any other, in, any other process mining project, Guys, please, one thing I want you to understand, and remember, first, if you bike way, how much, OK, forget about it, don't forget it, Do it. But the second thing is the biggest challenge in process mining really is to get the right data and to get the data right. And there are many different ways how to do that.

And the way we do it, you know, we realized that the data we need for process money is distributed and scattered across different systems.

And even if you think about a system like sap, for example, you know, one of the biggest, the biggest or ERP system, we are talking about 170 to 200,000 database tables.

So, if you want to have the data for one process, oni, like, like, purchase to pay, you need to reach 50 or 60 of the tables, You need to transform the data. So you really need to put it together. You need to create your eventual with your case study, your time-stamps and your activities. And only after you have the date, after you have your event blocks, you can load it into the process, my application.

And the system will visualize the process of social networks or kind of KPIs and statistics, But again, you know, one of the highlights of our platform really is that this entire, let's call it animal, really just a UI part of process mining platform.

So, we have connectors where we connect directly to the data source. The whole data transformation is part of our platform. We have an in memory database we use for that.

Then, we can visualize the process from all the different perspectives, aspects, and so on. And when we look at our application architecture in a simplified way, you know, we have all connectors, or conductors consists of the reader where we connect natively to all different applications, including ASAP.

We have our transformers. This is where we built the event logs and went, then we loaded the data into an application.

Based on the Dashboard, we have pre configured an application one, It's a template based on our experience, from some hundreds of projects, and typically we can answer between 100, I'd say, 80 and 99.9, of all questions with application one only. If not, we can easily extend it.

And we can customize it.

And by the way, this whole, this whole thing works on premise or in the cloud. So there's no need to go to the cloud if you don't want it. We have a lot of customers.

You know, we work for some, let's say, government offices and, and some very sensitive companies are on data, data, privacy. So they don't want to see the data in the cloud. So we completely support on premise. And from a functional functional perspective, there's no difference at all.

So, that's the way we are dealing with data.

And doing this data transformation process, we can apply all kinds of business rules. Like, for example, the very simple would probably is the SLA violation, you know, we want this customer actually expected that. An invoice will be paid on that, it will be processed within 30 days.

And, Hey, it's great. Look at this, you know, we, we're in the third quarter.

So, we really managed to bring down the number of SLA by donations from 503 to 130, and that's that's fantastic.

Well, yeah, it is, it's good, but, it's a really perfect because perfect would be no SLA violations at all.

And how can we cope with this problem?

What, the, the, the way we do that? Again? Is, we use the full power of our high bout mission platform.

Let's go back to one of the slides and, I was, you know, we have something we call the action center. It's over here and, what we did with our latest release was within two, to connect our process mining platform, again, to the action center.

And, what does it do? What?

Let me show you.

And, you know, you have seen this dashboard, and it's great that we have only 130 SLA vibrations in the third quarter. But, still, 130, 132 many. So, probably, in the future, we will define other texts. You know different routes, like, for example, we have an invoice that is viewed the day after tomorrow.

And if it's not approved today, then the system will create a tag, like possible, or potential SLA violation.

And based on this tech, we will create an action in our action center. So I'm actually actually can be, it's a lack of workflow, so people will be notified, like, hey, Rudi, there's a, there's an invoice, you, in two days. Please make sure you approve it today. And if I don't respond, because I'm in the hospital.

So the system will automatically use our RPA functionality either to, let's say, to reassign this Envoys to none to another person.

By increasing the priority by sending me a short message or whatsapp or we can utilize the full power of RPA platform. Basically, for example, to automatically pay this invoice. If it's a trusted supplier, we work before we go to with before because you know the last thing we want to do really is to to disappoint or partners. So we make sure we pay this invoice. In time, and if there's a problem we can still fix afterwards.

But we really, you know, we don't we use process wanting to discover RPA potential, We discover inefficiencies, we discover stuff like risk.

Of course, we use process modeling for monitoring, but we also use process money, in combination with our RPA platform, with the Action Center, to actively monitor or control processes, really two, to drive them, make sure that processes are proved constantly over time, and we get better and better.

Know, I could easily continue for hours like this, because we are, We have plenty of stuff here. Maybe one thing, I would really like to show you, very, very quickly, Because I touched on this, you know, how we connect directly to the data.

So what we have seen so far, is only the front end of our application that is available for every single user. But I'm logged in as a designer, or some admin. So that's the reason why I have this Duchess of took the pills.

And if I click on it, then actually I will get two more, three more tabs and two more columns.

So, we have data, we have dashboards, and we have to process crop, because the focus right now based on the process crop.

So that's the platform or that's a level where we connect to data sources. So, if I would like to connect life to a new data source, all I need to do is I right click and I told the system that I have a new table, I would like to use.

Of course, we can upload data, we have. Any format, it can be. It can be an exit file. It could be in a CSV file. It really doesn't matter.

Um the better way to do that as a connection string. So when I go for connection string. I get a new empty table. It's empty. I double click Y, right click, and I say, Edit.

I get this little box, It's not really fancy, but when I go to the head, I think it's a long list of connectors for read us, we have out of the box, You know, we can connect directly to an Excel file on the server or the client. Wherever they find the sheet to find the first column first row and read in the data directory without any staging or database in-between, we can connect to any file system.

He doesn't matter. We can, we can use ODBC connection to connect to any ODP to any relation database.

We have a native native connection to sap, not only for Ha ha, ha, nice, easy.

We can connect directly to the ECC system using the capability of some of our partners where we connect directly to SEC to ECC system. We defined the table we define, you know, what kind of data we're we need, the tables, the fields, airbrushing, and we can reach from the directly from the data.

We can join data no matter of discourse. Or lecture on the right.

show an image on one of the unique, joined the data with union tables, ..., Chazen, e-mail, Know, really, whatever you wanted to connect directly to the source, No staging area and no database transformation, no scripting, connect to the data, Transform the data within the platform. And you can even define data sources as live data.

Guys, you know, when we're talking about real-time monitoring, well, first of all, in process mining, there is no real time. Why not? Because first, we need to see the digital footprint. We need to transform the data, and then we can see it in a dashboard. So, if anyone is trying to sell you a real time process monitoring, process money, sent them home, doesn't work. But with the live data, we come pretty close to that.

Because every time you open the dashboard, you open the application, will pull the data, or read the data directly from the source, transform it, and show it on the screen. Of course, there should be careful. You know, what, you define this live data. Don't do it with the change documents, sap. They are better ways how we can, you know, really enabled monitoring here.

But, this is a great way, really, to come as close to real time as possible, OK?

Let me delete it again, you know, the other, the other tabs. This is where we defined the dashboards. You know, we don't have the time to build some new dashboards, but I would love to show it to you one day.

And because it focuses on the ..., and where, here, we can decide what is our Case ID, what is the activity, What time, the time-stamp.

And you can really create and look at the process, you can create whatever you want. And you can look at the processes from every perspective you like, And we even have some advanced, crazy stuff. Crazy. But you know, for example, if you would like to see this process vertically or horizontally, you can simply switch, apply the change in you, see the process from left to right.

Let's go back to vertically. It fits better into this dashboard.

OK, so what I wanted to show you in this way is to know, I think really, I really believe that we have the most comprehends, the most flexible solution out there.

Not only for process mining, but really embedded in our entire high power automation platform, where it really makes sense to me, to discover processes before you automate, and you can monitor on you can control the processes, and you really have the full flexibility.

And talking about flexibility, the last thing before we switch to two questions, know, let's show you one nice example.

You know, over here, let us say, we take only the cases where we have this checked, and I'm proof, that's the sort of a fast tracking my process. And what I discovered is that for some reason, I see a loop in my fostering, which is quite strange, loop, and the fast track that shouldn't be there. Well, so what's coming up in a one click, of course, will bring up the 12 cases invariant 25, OK, Marine 25. So what's wrong here?

You know, let's, let's focus on Varian 25 for a moment to better understand what the problem is.

First of all, I see you know that we F a.

We received the invoice then again, the thought process to invoice, and then after four hours we come to the first check their pool, which is perfectly fine.

But then we have seven days, in average thought for a second, for the second check, the proof, which shouldn't be there, then even if I unroll the loop to better understand what's happening here or days for the first 1 but 7 days for the second one.

So, if I want to better understand where this product comes from, of course, I can, I can go to supply us, and what about what maybe we should talk to a first glance.

But, if I really want to understand the source of my problem, I need to know to shift the few, Like, you know, like they did in the hospital with my X ray. They actually took this way from from different angles, and we can do the same here.

So, if I go to dimensions, I can actually look at the same process from different perspectives from different dimensions.

Of course, you know, activity is only one. So, again, when we add complexity, we see the loop, but I can switch, for example, for something like country.

Now, we see the same process, but now visualized based on the, on the countries.

And we see that, oh, it looks like we have a problem between Canada and Netherlands because it looks like the bottleneck is over here. So, you know, let's switch to the team.

So, let's coming up, OK, so it looks like the problem isn't operations, marketing, And we can even go one level deeper, really, to the lowest level of users when it's a big picture. So, let me move it around a little bit.

So, now, we see, you know, we're, in the third quarter, were fully implemented RPA. So, we see the different instance of our scanning paths to different instance of our processing boats, And we have only a few people involved in the process.

But, well, it looks like Kelsey Hoover might be a problem in this process.

Let me guess she's based in Atlanta.

She works for operations, marketing, and if you don't believe me, you know, again, let us go to the to the ditches, We can see, it's disordered on time, we take the long lasting case.

Oh my gosh, look at this. So, basically, that this case was, or this invoice was finished after 18 minutes, but then Kathy Hoover took over and she delayed the whole thing by 24 days.

What we still paid on time, but that's not the point we need to understand you know how we can have Casey to improve this. Maybe she needs to botch in its attendant bot to hyper or basically probably shouldn't.

We should tell her that we don't need a second checked and approved OK, guys Basically does it you know we have much more to show.

I would love to go to every single detail But we don't have the time today. So if you want to know more You'll find me on LinkedIn. Please reach out to me And I will be happy to accommodate another demo and really answer your individual questions.

So I know that we got some questions here.

Really great presentation, we really appreciate it. I'm gonna ask you to click that stopped showing screen button so that the audience can see our video feeds, their screens. Yeah, thank you, First of all, thank you, man. You're just fell last week. And you're here with us this week, I can look at that, and I can imagine that there is some post surgery, pain, and drugs that happening, and I hope that you're feeling OK, it looks on your way to recovery. Thanks for really being here, this was a fascinating, deep dive on your process mining, RPA, and all these different capabilities, as you mentioned, related to the product topic of hyper automation. So, because of our audience, and we have only a couple of minutes here, I want to thank Grace Wong, who answer some of the questions directly to the audience, some of the tactical questions that came up during our presentation.

I want to ask a bit more of a strategic question related to the, to the audience that we have. Lot of business transformation, operational excellence, leaders, work in our organizations across around the world.

What is the best way that you find that a company like UI Path can partner with those internal consultants to the business, if you will? Those operational excellence leaders who really understand from how the business works. And, what can they do to better partner with you, to get your solutions faster, to get you a better solution, chance? What does that look like?

Know, we, we asked UI Path, We have a lot of experience with over 7000 customers.

We have a lot of partners, all our partners are getting trained in process mining, and they are really combining all the different tools we have in our platform. So if you are, let's say if you're a specialist or leader in business transformation and operational excellence, you know that diagnosis, again, is only part of the job. You need to understand what the problem is. But you need to think about you know, how the process can be really improved. And again there are many different ways but RP is a great one. So I would really encourage you to reach out to us most likely your company. Or it may be that your company is already one of our customers, or maybe you know, that you're already using RPA. If you are, that UI path, process mining. And ... really is the best combination out there. Because we have other connect us out of the box with the integration with our, with our bots out of the box. So you will not only see what their challenges are. But after we've implemented RPA, you can immediately see the impact and the success of RPA on your processes.

Let's talk, that's, that's the best of the best much for that. Our time is up for the session, but we really appreciate you from Frankfurt, Germany, to the world, sharing our knowledge or expertise deep insights into this very important subject. Thank you for taking the time to share that with us.

Great pleasure. Thank you for having me and guys, never forget where if you If you ride your bicycle. That's right. That's right, well then. Thank you very much.

Thank you guys. Take care. Bye bye.

Take care ladies and gentlemen. That concludes this segment. At the top of the hour we're going to be welcoming our next leader who will be discussing mining, modeling or management. Where should you place your back? And we have a world renowned author, performance and business coach on Mark McGregor sharing his insights over 30 years of experience on business, process management working with technology cross industry organizations. A wealth of knowledge and wisdom that Mark is going to share with us at the top of the hour. Please, if as you close the session, there's always an opportunity for you to provide feedback on this and any other sessions on the goto Webinar interface, and we look forward to seeing you at the top of the hour with Mark .... Thank you.