Session Information:

End-to-end process optimization and robotics process automation (RPA): creating value from design to execution

RPA is one of many levers for improving processes and operational performance. Yet, it often lives its own life in large organizations: from the organizational set-up and resources to identifying and capturing opportunities for process optimization. Substantial value can be created from integrating RPA into end-to-end process management; and leveraging it fully to enhance customer experience and improve efficiency.

Takeaways:

• How RPA relates to other improvement levers; when to use RPA; and when not to
• What is the optimal organizational design; and how to ensure RPA is well integrated in process management
• How to run RPA discovery; and identify the levers to improve end-to-end processes
• What value can be captured when doing it right

Session Transcript:

okay folks 2:15 so if you are killing a star dear like to choose to get some attention and he will be our speaker today and we actually share a lot of background very similar background both from Eastern.

Europe both beautiful I fly from Boston flight from Amsterdam so that's that's the difference we have both two kids it's a great guy presentation the body process automation how to get more out of it gotta get more out to be in the context of end to end process management and how to really take it in very structured way from design to execution love-bomb boring you much I just want to take 30 seconds to introduce Angie who we learn because I think the context is also important for our RBA threats very global man we present in around 45 geographies roughly 15 of them mostly Europe Australia and Asia is from universal banking 45 of them including the feeling was that I mentioned our coastal back you know me so we have pretty.

when the truly global footprint which especially we start something new like rpms well be quite a change when she started with robotics quite some time ago we already have a history study she's essentially 2007 when I am she merged with another large bank in the Netherlands post bank and then make you some scripts to help move around 1 million customers we estimated that three heaven if we wouldn't eat that that would've actually have to invest a lot in both fps and IV so this is what we consider the financial RP energy pretty much at the same time we started with enthusing computing in Poland.

when a local team does follow me an idea that they can came up with macros if you live that were installing the bestop of every employee to help me routine routine tasks and simple computations then it took quite some time until we decide what to do in RB 2016-2017 when we really build it as a global program this is when you move from two countries to running our APA on global scale some of the main cornerstones of our water and then we're going to go into detail speed of that first and foremost we do not see RBA as an improvement leaver on its own we already can see our PA as one of the leaders to improve our processes and second and this was quite as with every new language it takes time to network to learn when I'm speaking the same process
the approach that we introduced a year ago.

Every RBA opportunity starts with process discovery and to and process discovery where we look for improvement opportunities not just for PA we have global framework we have developed it in
her office and we're both asking and supporting countries and Laurie needs to implement it because this is saving them lots of time and again is helping us maximize professor Glossop not least we have to introduce proper governance with a very clear role for the global CEO a week for the global center of expertise as what was the locus is always in the video let me spend a minute on that EDM and eliminate these guys move manage.

we see this coming as a call and ideally in the right thickness what does that mean when we look at the network process we always first try to eliminate non value-adding activities think very lean think various quick wings think of it as things that you can do without the products or without automation then if we can digitize something we if we can automate something with structure or tidy solutions that's always the preferred method for us so if we can develop a fancy front-end application for cloth for clients to open up any banking account let's of course the best option however oftentimes this is not this is not feasible because of limitations of time or IT resources or projects so then if we cannot develop a structure ideal solution.

Nine months that we're on guys then when we have clean the process what we have eliminated and we have decided what to implement what to improve in structural I didn't want to advertise then the activities which are still left to be performed manually that we still question ourselves truly perform them when we performing them now which should be moving somewhere else this is what we mean is about moving steps of processes within same organizational unit across beams in the same location or even two different locations billing initial facility or outsourcing and then we repeat this again couple of years later we look at the same end-to-end process and for exact copies.

It hasn't always work like that in practice we may have reasons to deviate but this is our thinking this is our philosophy this is the best way that we believe in when we talk about and this is a radical shift about what we're still doing the couple of years ago a couple of years ago our local units were
coming up with bottom-up technical tactical initiatives throughout our society so they'll be looking at the process that we don't see in our menu ability don't talk to IP there will be no budget in them and then saying okay let's become a product we're still doing that but less less state now we are at the identifying RDA opportunities from an end-to-end process perspective and when questioning and challenging ourselves what are the right things to our PA and/or are also things that we can move
speaking the same process now a language that's very important that's very important for any institution that's even more important to a global institution 55,000 people spending across different continents this is our definition of an end-to-end process.

A process is what we do speak when systems to deliver meaningful welcome to our customers and businesses if you aren't putting in a different way these means completely satisfying a customer need everything that we do to completely satisfy customer needs where the customers are the external customers and then the internal customers are our state cars why do we do that because anything short of completely satisfying the customer name is actually not executing an end-to-end process what is important if you have a customer that applies for mortgage and you only collect their documents and you tell them well yes mainly I'm going to give you the loan but you don't disperse the money you have a new field any need thus you cannot execute it the entire process the 30 wouldn't first work is welcome because now we use our global process catalog this is on page our end-to-end process catalog

Where by nineteen and qhorin supporting processes we describe body roughly content with the end-to-end process that we have energy when we have this in place then we can start mapping our RPA opportunities to be same and to end processes a year ago if you ask any of our local units to tell you what their local RP a backlog would be on them.

I wanna be they don't come up with their own language with you know definitions of technical opportunities that couldn't be linked to any end to end process that's not the case anymore for every single opportunity were able to assign it to an internal process and actually because we use the same language were able to compare how different countries ruggedized exactly the same end to end process so do we see the same opportunity in other units as well do we see three units working on Robert I see exactly the same piece of the process that has given us big boost in terms of bone scan speed as shown in example core processes that's an example of supporting processes again.

We had units that work more in core we have others that work more supporting but we use exactly the same language and classification of our RPA starts with mental process discovery a year ago when we would launch an RPA development who thought with the intake so typically somebody in operations repeat them would approach our local CEO in say hey I have this great idea to aromatize that you know time delay then would apply some criteria amount of protein in the leaders for the criterion it's basic atomic volumes it's about frequency it's about the complexity or the simplicity to task to notice then a DSS and appositive would implement so we develop the script to test them and then deploy them and then would sustain them with our infrastructure risk in there is a story ins in terms of making sure that we meet our law.

I suppose what we introduced the year long was the discovery phase and the way we start now every part be opportunity identification is with the discovery phase and unsurprising we run discoveries well from that during process perspective how can we do that first of all we need to have a couple of prerequisites in place then we form a three step approach to running discovery
and then we always want to get a very tangible value driven power to the company the prerequisites when we look at the end-to-end process won't be very clear and we do this in the preparation to discovery will be very clear about which of the locations that we'll need to this how long is our RPS can be going to Lots which will be end-to-end processes remain even the broad scope of them that we want to add you get then every end-to-end process.

That we run discovery for we somehow Gordon see it's not sufficient to just run interviews with all stakeholders and participants but you want at least at least short on the shop floor demo visits then we still talk to the SMEs to the subject matter experts and then we take this into prioritizing our opportunities the outcome is very clear definition of water the process is to be improved and again which part is to be improved by robotics and which parts can be improved in other way then what the savings or the benefits our process and then in terms of time and efforts how we can prioritize these or entities you can imagine again in a very simplified fashion the output of a discovery to look something like that you have the internal processes you have the scenarios in the processes and then you have the number of opportunities identified earlier running such a discovery is not very long typically takes a day to three days with planet for the most complex processes.

Main is going to take five days most of these most of the discoveries that we plan so far take two to three days it's time worked well worth investing for us number for our global arena structure we have a global infrastructure that we are now applying to all our units we I use automation anywhere.

it's always gonna vendor so in terms of our being it's the a to name we have an internally developed arcade - it's worth it's called rubber platter there was actually about the important that started with robotics ten years ago in us and we're now trying to be fine very clearly the use cases for anywhere we introduced last year a broad resource system so that we can address lots of risks and compliance issue that we have so now every user can request their own account in order to generate their passwords on their own through the robot resource systems so we eliminated all the risks in terms of compliance and fraud and these infrastructure is now used across energy company.

I mentioned risks and so these are the global risks that we need to comply with one of them is some cybercrime heading of robots unauthorized changes to instructions for scripts robots before introducing the road resource system actually an ops employee gets to set up the passwords for odds this is not a case anymore Prabhat to generate their own passwords and that's completely we meet these risk.

Requirements now on a global scale again why are we doing that globally because once we managed to mitigate the risks on a global level non-work unit needs to learn on their own so this is the support that we're providing is silly as well among others let me list all of them this curve Ocoee is organized there is the MOBOTIX squat which sparked my team the loan process maybe they say in Amsterdam this is just for people and then we have two squads of ie custodians and scripters they said in Poland which is also low-cost location the total number of people he needs to is 12 so.

we thought do we have 15 16 people for robotics in the global Siri what we do is then AG start with the Estonians and scripters the custodians from being restructure so all the robot accounts all the risk mitigations that's what they do the scripters that's our global scripting unit we can send scripters to all local units that need to become BOTS and they do it following exactly the same approach using exactly the same - and leveraging as well scripts from the different countries.

Where we have implemented similar use cases and then what we do in Amsterdam is really one thing it's the global greens it's the benefits it's the use cases it's making sure that healthy run RBA is fully in line with our process management approach and then we still support all the logo co easy residues we do that in the following way we have our casino needs we have around 70 than now so in all the local units as well as in all those versus the local units are really in charge of developing the scripts in the long wheel developing the script and then maintaining the script as well.

Using beads with this component or we have a couple of other options as well that I'm going to show himself and then there are some the ones that still do the intake in the local stakeholders in terms of identifying our opportunities Pokemon and making sure that they comply with the global framework and our global.

We have been experimenting with different implementation models the joy still without but I think we're really shifting more from you know one to get it close before let me just briefly go through the four moves that we have in place the first one is what I call the operator morning you take a subject matter expert from say operations worried that you need somebody who has an interest in your voice not an IT person though you basically train them to develop scripts there can be benefits so they we know their stuff.

They know the process very well and obviously if they want to start developing scripted scripts they are willing to do so so there's a great deal of energy in exam the changes are typically these people do lots of other things as well capacity is an issue speed is an issue sometimes training and experience can be nice respond second part of the consultant model we take the essentials and the cognizance of this world and we use with their scripters to help develop scripts either in a loan we meet or ideally dissertation.

Globally as well the benefit are these are trees guys they can bring in scripting knowledge from other clients as well they may help us with broaching an RBA opportunity in a different way maybe in a way that we have dorba they're expensive and they don't let us build a sustainable model for well RBA.

Inside ammaji we're still working with them but we're working less and less the third model having scripting capacity in Belarus EoE blasio we intentionally so that would be another role that they would perform educating local scripters in Syria as well the benefit is well they really become part of the organization the main challenges that require speak up from the investment and then if you don't get a solution the amount for our basic reading in the local unit you really need to find a way to deploy these guys globally.

Which is a bit more difficult than if you do it on a global scale straight away which is the fourth model which is our globe winters in Poland that we are now deploying across multiple different unit starting and you can see that we can even you know the same unique still working with different models Belgium is great examples of Belgium still has the operator model we can sow the bottom and now the global sleeping but we're clearly shifting more and more towards three more what are the results so far for our three years of global implementation and then looking at what the Netherlands in Poland hit done in the years before we have a release the right for convict empty community basis we have introduced complete 7300 both runners and we have achieved.

Savings words roughly ten video for bank like Angie this is still scratching the surface however it's been a good start and a stir that already gives us crime syndicates what's next I didn't wear this in a couple of the other automation and rpm presentations over these the best as well you'll deeply start with our PA then you move to cognitive RBA then you move to chat box and then artificial intelligence.

Again they don't need to come in this sequence so you may also do this imperative of course but this is if you will be evolution we're still someone here we already know how to work with traditional RBA we still need to grow the scale speed to capture the full benefits me my estimate is that we probably captured around one quarter maybe one third of the benefits that we can but just using simple traditional are people moving to cognitive we've been running proof of concept over the last 20 year.

Again with automation anywhere in a couple of other tenders we're using IQ bot and we find some very good use cases for cognitive are being Bank as well our priority for this year is to increase the scope of the use cases and really move to implementation so that we start capturing benefits through cognitive.

Happy as well we know it can work that we're already convinced about so it's done we captured some balance in terms of chatbox and artificial intelligence we'll still pretty much in the stage where we playing with it so what we'll learn experiment and we're running become the pilots we stay close to different vendors from new opening up but is he still not something that we can really commercialize and get you know savings to a customer vampire this is where we are in this is how we're thinking of the LEAs the use.

It this is what I wanted to tell you Bob it's a mix of real savings and cost avoidance as well and I witnessed a discussion in the top of the breakout so how do you actually show or how do you are there to demonstrate that cost avoidance is really outcome of the work that you do we still have the benefit that we're growing very fast in many of our minds we have lines of business let me give a couple of examples mortgages in Australia.

Consumer banking still invest in you know when we grow at 20 30 % in a year and if we manage to achieve that growth without carrying the digital people we're not so fast about you know that from our PA or does this come also from something else so when the benefit is there who don t know jealous in competitiveness to who's the main driver so this is going to measure we really track business growth and if we're meant to achieve this business world without cost investments then we consider this to be really bad however we also have some real savings so yes we make people
leave without the bank if not the local unit at least the line of business in which they are working and we may really.

You know organizations evolve sometimes they go from broad-based benefits to P&L benefits to by resource analytics of what happened right so one question I have for you is having balanced ROI with the people you know the worry the fear factor you know the change management side of things so you have to deliver money but then you also have to figure out how to take back gee let me uh that means you go first and foremost remember position process management and then when process improvements as a cost-cutting issue we clearly say we can you say that we want to achieve couple benefits.

One is improving for a period but you can improve productivity by either reinvesting to achieve growth or by a real cost reduction we're also very explicit in terms of the customer satisfaction and employee satisfaction that we want to achieve for our period we're still not at the point where we would say you know we know that this are being opportunity or if we introduce this Robert it's going to increase customer satisfaction by X when you measure customer satisfaction at the
level of the end-to-end processes.

So the mortgage leads to cash example so this is where we measure the productivity increase as well as the customer satisfaction increase as well as the employee satisfaction increase however for monix were still able to say well we know that robot eyes in that part of the process is going to give us that much capacity reinvested or release and this we track due to the fear of employees once you very afraid ourselves in the beginning that people might not perceive our being the right way that fear is pretty much not there I mean we jump pretty quickly across the change if you know convincing people that if you introduce a lot the job will be gone and they will be out of job themselves so we either reinvest in the capacity perhaps or in case in the case of true here called sayings these would have been processes that would have improved anyway so maybe we're lucky but we don't care for much of an end for each other's day I think we managed to create a pretty good case we change around avoids these what triggers so you said you started really doing a lot of local automation then you backed out so let's.

Take a little more ballistic process you drive through your new what what trigger that changed what broke the change in your approach that's a great question 100 I think we just started acting when globally energy overall until two years ago I would say we're you know a group banks now want to be banking rode through and we're doing this not just process management and robotics we're doing this for multiple initiatives as well so I just culture is very much prettier one lots of things happen you know bottom-up and with the entrepreneurship and innovation of local units which is great we don't want to lose that true however we also tend to face situations where you know three or five law go back to work on exactly the same thing at st. all the time that's not the right answer that's not you know the right way to spend our time.

This is why as well we want to create a case for change by saying if we want to do things in a global way we need to create with the right global ways but then if we do things globally then we don't replicate and only be the same thing many times you know it's very difficult it's not complete but far enough I think we've made good progress or we are making good progress but we are not a widow there yet it will take some years with a thirteen hundred provides you have a production whose
responsibility is the monitoring.

Major care of those robots is if the local centre of excellence bowls and excellence we talk about the monitoring and yep the vote see only so much doesn't doesn't mean the once we will track they use it to the global licensees we track the benefits however the maintenance is done in terms of making sure that the script run by the local series and then in terms of implementing and running the infrastructure so private cloud the robot account system the risk mitigation factors this is the goal behind the story.

You're doing this all these different countries like do you maintain the actuals static architecture for all your value chains that these robotics actually adhere to or is everybody kind of running their own process and and Poland vs Italy or Oscar the question basically is do we really have global processes in place no we don't know no we don't yet so that'll be my assessment but we're designing several and we commend is that a realistic we think it's very realistic to get across ing not just one set up but I saw one that you call it the order to market or something like that yeah so we're working we can't be some designer already last year three vocal processes on asleep again at any bank we need to cash in because you know that you need to catch so we already hear this process on paper.

What we're really doing now is we're preparing for implementation units we have that in too many if not for tea at least in some units by every year so we think it's very realistic if those around to his work to get where we are however it is very realistic still for this kind of doing doesn't mean that we're going to get how many percent exactly the same movie to cash processing over between countries working out regulations to differs most we think we have different in our chairs in different countries we may still be having some different customer requirements because who needs but even for mortgages which is pretty much you know the most complex with their business in banking.

We think that we can be seventy percent say do something you have to give up in terms of understanding overall performance of it you're having different levels of performance to do that you're just going to understand that together we actually think that doing that is going to help us monitor and manage performance in that way yeah just be doing that in the same way across Angie please do you have like a certain threshold or desirable number of exceptions allowed but it would say will automate this verse that was yeah we do no it's not it's not one single number so it might differ by processor like that process but we do yeah so we also can and I was shown this here's one so you get automation anywhere the wrong platform.

One way we considered to be this thing to distinction between duties when you don't mention anywhere for unattended use and they're basically want to minimize exceptions and withdraw a platform which never seem to be using it with you know with human steady mode and there we can allow ourselves a slightly great deal of exemptions but we typically want to stick though we don't seem to be these numbers for 70 20 times exceptions so many good things start from the don't we have to see your ears we calling over call the energy switch you think of your Cal form to customers are banks sound.

Easily so we really want to be a lot brand for financial institution for our customers and starting from our CEO we communicate and incentivize we support Community Network one thing is which is taken across customs so we talk a lot to take in multiple different ways as we were very direct bank in Germany so we don't have any tranches being direct allows us as well to research customer medium because my main points is an easier way but even in the countries were still very shaken branches we really take the opportunity together because there is a doctor and then now it's also an hour you know values you know ways so we have an article behavior and things that yet for customers you saw in the video as well is really something that every new we learn the basic inventors they'll jump.