Session Information:

Real-World Use Cases of Intelligent Automation Part 2 - Classic examples of IA across different verticals

Join this webinar as we talk about real-world use cases of Intelligent Automation (IA) across key industry verticals, including Banking and Finance, Logistics and Transportation, Manufacturing, Insurance, and Healthcare.

You will learn how Intelligent automation connects processes, humans, and technologies to drive revenue, reduce costs, mitigate risks, and boost growth while delivering superior customer experiences.

We will showcase how the application of advanced technology tools (such as computer vision, machine learning, natural language processing, and others) to a complex process can automate it from beginning to end, for purposes of reducing human effort and increasing efficiency.

In this session, gather insights on:

• Common business challenges post COVID-19 pandemic.
• Industry-specific IA use cases.
• How Nividous IA platform can deliver greater ROI.

Session Transcript:

Fantastic. Thank you for joining us.

Ladies and gentlemen, shuttle has over 25 years of experience in the process, automation space, and brings tremendous knowledge to help clients navigate the digital transformation journey.

He plays a leading role in steering innovation for the ..., one of a client of hyper automation platform.

This vision is helped snippet is get recognized as high profile for a driving digital transformation throughout intelligence automation. Please forgive me for that. Apparently, one of my bulbs just decided to check out for a second.

Thought on that note, I'm going to pass you over to wrestle and shuttle, over to you.

Thank you very much.

So, thank you, Brian, for the introduction, and thank you, or being on this webinar with us today. Looking forward to walk you through some of the insights that we have in the intelligent automation space.

And to start off with that, I'd like to actually introduce, uh, what we are as a company, and a brief introduction into intelligent dog.

So, Brian, let me know if you're able to see my slides now.

Brian, are you able to see my webinar slide?

Yes, we can see the slides. And testing. Thank you very much. So, thank you, folks.

And as Brian mentioned, I am the co-founder far off of it, is, be a focus on intelligent automation in your product suite.

Now focuses on this, but before I move into that, I'd like to give a little introduction about our background, and, uh, the vividness Overview, as such.

We are 10 year old company doing great work weaving, growing very rapidly, nano customers, with a very high success rate.

Mean, percipient of many industry awards in this space and the reason we have been able to do that is because our focus, he's around one of the fastest growing segments today, which is business automation.

Business automation actually comprises business automation or intelligent automation are used interchangeably and robotic process automation is one of the components off.

Such an intelligent automation platform.

Business process automation forms.

The second part, and into artificial intelligence, is it sort of show, are the focus. Today will be to walk you through how each of these components play a very important role in your process automation or your business automation journey.

We cater to customers across all different segments: banking, financial services, manufacturing, logistics, insurance, healthcare, life sciences, ITV's, PO to name just a few.

We also acknowledge by analysts leading analysts, like Gartner has covered us as one of the top products worldwide.

Ah, and foresters cover some of our clients. Customers seen some very interesting research papers. Every group has been covering as regularly, as well as the fastest growing products in this space, and also maxes out specifically as a performer.

and this has happened multiple years in a row.

This brings me to a port.

So, before we even start, I know, let us understand how much you are invested currently in your automation journey. So, are you currently running away from automation in your business?

We see that, you know, than the lower customers find it challenging.

They know that this is required is important, but they, for some reason, find that it is very hard to achieve or very difficult to target or start with chill.

That's when you run this full course, the polls are open.

We'd like to hear your feedback Like to understand what you feel about your personal experience in your enterprise Amanda's journey.

Answers are coming in the show.

Yeah.

We'll keep it open for about 10 more seconds to give people a chance to join it, and then we will show you the responses.

There is an overwhelming favorite at the moment.

So, I'm going to close that now and share responses there. We have a treadmill.

So, um, helped me through this, so, Brian, perhaps I'm not able to see them.

67% of the audience has said they're currently running it in production, and 17% have a habit in the pilot, or proof of concept phase, and matched by another 17% that are planning to explore into the next 12 months.

Fantastic.

So, this is consistent with what we are seeing, so we do see that automation is the need of VR.

We do see that folks are either already on their journey, they've adopted the good part of it.

Maybe they're already, um, mature.

Many of them are just starting off on their journey, and some of them are still thinking through it, but certainly, not many that we have seen are running away from automation. They're embracing automation.

So, this Board was designed specifically to, to bring out any outliers and NBC.

There are none in every poll we run this, We have seen that most people are overwhelmingly will then lead, saying that they are, they have adopted, or on their journey already.

So, great tool, Let's recap what we talked about. Sorry, was there a question?

Nope, nope, nothing at all, when I was just wondering, if you can, during the course of this discussion, share a couple of case studies to determine Stripe. How your platform is help clients across the globe.

Absolutely, great, great point over there.

And, uh, what better way to illustrate what we mean by intelligent automation by, other than, you know, I'll walk into a case study.

So, before I even start on, on a case study, you know, what is intelligent automation, maybe a quick recap from the last time around.

Intelligent automation means different things to different people and specifically in different industries.

But basically, the application of technology tools, and specifically computer vision, machine learning forms, predominantly a good part of that. There, you know, the intelligence comes from where you are giving the cognitive flavor to your automation.

It also includes the natural language processing, where you actually know, how the cognitive ability to interpret things like a human. And then add to it.

Any kind of manual activity that a human, so lift and shift operation, which comes in with RPA, which reduces your effort. And then, also ties it all together.

two a chain of no triggered process automation. So basically, one task finishes and the next one takes over and so on and so forth, which is really what your enterprises, so that's really in a simple way, intelligent automation, adding intelligence to a normal automation that you would try to achieve in your business.

And recently intelligent automation is also given different terms. Hyper automation is another term that is used frequently by analysts such as Gawker.

So, uh, pretty much all of them point to the same thing and then basically trying to get unbeatable returns on your investment at a multiplier level.

So how do you achieve this?

Like I said, there are three real parts that, that, know, complete the whole occur, intelligent automation. This is our opinion.

and this is largely what industry is talking about now.

So the robotic process automation we have seen has been around for awhile now is being tried and tested. People have achieved a lot of results. Interesting Ottawa results to adopting it. But they become siloed automation.

So, ..., we hear a lot of customers saying, hey, I'm done, good amount of RPA work, but, you know, I don't know what to do on the sub acute, low hanging fruit. But I wanted to do my enterprise automation. So, how do they achieve that?

So, Seiler automation is what good, you know, be achieved through robotic process automation. But when you want to add human cognitive intelligence to a bot, is where artificial intelligence comes in. And then, if you were to add human perspective there, lot of decision making has to necessarily be done by humans, that you need to approve certain things, where you need to handle certain exceptions, where human comes into picture.

So the BPM, the business process management comes to BPM, which orchestrates all the discrete tasks, and those tasks could be human, tasks, could be robotic tasks, or it could be integration, does your organization.

So, entire gamut of end to end automation is what we call intelligent automation.

So, if you are running, or planning to use, uh, so this, this pretty much brings us to the next poll before we move on to the case studies.

You know, the poll that we'd like to present is an interesting one here.

You know, if you're running or planning to use intelligent automation, which of the following challenges should it support, OK, so there are multiple choices here.

If you are running or planning to use intelligent automation, which of the following challenges shouldn't support?

So, I talked about intelligent automation, give a little overview from our perspective.

What is your perspective on the challenges that it should be able to manage?

So, random, I'm hoping that we already have people sending in their responses.

We do, indeed. It's pretty even Stevens' at the moment, actually.

Uh, well, it's spreading across everything.

I'm going to leave this open for about another NaN, so I get those answers in, and then we will share that.

Um, OK, let me close that and share the results, So we've got those. So, we have manual data management, which obviously includes data mapping, entry, extraction, et cetera, 80% of you are doing that.

Improving the integration capabilities of core systems, 20% of view of responded on that one. Optimizing existing business processes with automation, about 20%.

And obviously, using it in conjunction with other technologies like BPM, 40% of you have mentioned that, so trivial that, that's your answers across the piece. That, that is pretty interesting. So the last last answer itself is a, is a very interesting one, because it shows the transition that people have already started making social and RPA started about a few years back.

You heard a lot of the RPA vendors start saying that long live RPA, BPM is dead.

Many of you may have heard that terminology, the slogan out there But we understand that these are all, you know, to be to be working together. All of these technologies work together.

It's not one replacing the utter and you have already understood that by adopting and moving beyond from home.

You know, so let me call it this way. So, this is a bottom-up approach.

Traditionally, in the past, people used to automate enterprises by doing a top-down approach or a bottom-up approach.

Top down, of course, are long running problem programs which may take longer to get your auto lies, or no, you understand if you are going to achieve, successful or not.

So you may go to many, many months until you come to the end and then realize that there's some failures or a problem that you have to address and gets into a very long winded no, no rollout. So, that leads to challenges.

So, what we have seen recently is people going from bottom-up to no till, start with RPA, which allows you to do this, you know, plug in your own, give it more intelligence, then add other layers on top, and addressed smaller partial organization functions. And you do that, you can start using VPN.

So, what is interesting here is that all of these products are all of these approaches, involve processes, each of which can be managed, so you don't have to live out your best process and throw it out there. You can even start with something simple and keep improving over time. Whether it is an RPA process or a BPM process.

So, know, adopting it rather than resisting change, adopting change, is immediately, and that can be done easily. If you have a system, an intelligent automation system, that is a trial, then allows you to create new versions on the fly. And keep upgrading quickly.

Know, I'm not, we're not talking months, we're talking about the week, sometimes rollout new new versions in weeks and keep improving as you move up from from the bottom towards the top.

Just to call out a few.

A few requirements to this end.

Let me talk about our intelligent automation and how it addresses this requirement, So we have studio, for you, it's in development environment for where you would create and deploy robotic and BPM processes into what we call the control center.

This is the heart of our system, which controls everything, and also execute all the processes, and here's the BPM processes.

The RFP processes will be executed on botch, and smart boards are where your intelligent automation would run.

So like I said, BPM, the Human Tasks, Monitoring, and Management aspect, which gives you the end to end visibility of your enterprise, it's something that binds everything together through the control center.

Bob?

That brings me to certain case studies or Brian talked about earlier.

He was keen to understand how we have achieved this, and what are the different no industries in which we have done this? So, let me start off with a very, very interesting one.

It's in an automotive company.

A very large automotive, perhaps one of the largest volume companies.

We had a requirement or a cash back bonus to loyal customers, so a customer that is buying another car from the same vendor from the same automotive company would receive a cash back bonus for being a loyal customer. And this would happen to dealers. They were almost 5000 dealers across the country.

It was a huge volume requirement and it requires dealers to take and documents, you know, a lot of car related documents, invoice related documents of the new car, or car insurance information, registration information, all of those things.

And these have to be validated that pushback, yeah, to certain systems, to the automotive company.

From there, they were validated, uh, Virginia, again, a third party websites, government websites, and their internal systems and then auditors were reviewers.

And then the cash back was released.

So, so the bonus was really the contract.

So, these are several challenges as you can understand. Data extraction is a big problem. Then, also, having a lot of people in the back office to validate all of these things against all these different systems and third party websites is a challenge.

And the end to end SLAs were never met within time, ..., especially. They would have peaks and where, you know, towards the end of the month, they would have a lot of people clamoring to put in all their claims.

So, what we did was we implemented intelligent automation here which started with up the documents being ingested by dealers.

They were actually OCR, intelligent OCI was performed on those documents right at the point of injection. And there were some ability to correct them in case they were, they were wrong.

So, a lot of the data extraction was handled automatically, and a few corrections for bad documents or bad quality document was handled right by the dealer themselves and then fed into the the process flow.

The BPM came into picture, that it will trigger an entire process flow, and it will go to the first step, which would be triggering box. Now, bought to take this data, validate against a party different systems, internally, as well as external, government, and related websites, and all those rules would be validated against the data. Matching would be done, your first name, last name, what could be reversed in different places, so a lot of other intelligence mentors.

Then, finally, an auditor will be presented with this information to say, you know, you know, this may be going through or not, and some of them could also look, illustrates the processing of everything.

So, this give a lot of, you know, automation.

The team size was reduced to more than half, and the team was put to use in other areas of improvement.

They were the volume.

This was huge, one point eight million documents per month and and now, this was such a huge success, that they are replicating this for other processes.

Um, maybe three times the wall.

And 90% of enforcer are completely removed from this. Because the intelligent automation was taking care of everything that a detailed audit trail get the full visibility end to end. There were so many other advantages, just from, other than a net ROI, in terms of average savings.

First time, right?

You know, getting it done on time, SLAs, being matched, all of those were very, very important characteristics of the solution.

Another very interesting one, in fact, one for which our customer has actually filed an IP.

This isn't a logistics and logistics management company doing shipping of large containers from Europe.

And they had, you know, very skilled staff.

They had to do tedious verifications across multiple documents, starting with the bank LLCs, which had a lot of mandates.

And any good notice, you know, if you understand how these bankers, he's well they're completely text. There's no, there's no structured information. Unstructured text. There are all these mandates.

For all the different document requirements are presented and all of these have to be validated against different document.

So, bank elsie's triggers the process from the ERP system And uh then there will be bill of lading. There will be various other documents including finally invoice document. For shipping documents in-between NYS documented the very end that had to be presented to the customer, then everything was set up. So, different stages of the process required, complex document, understanding, and extraction. Some of them being Word documents, some of them Excel, some of them just scanned, some of them PDF.

And these documents had to be annotated.

So, whenever there are problems, that's the bot found, there were there were intelligent bots. So, you know, machine learning was used along with regular RPA bots and on process flow to do the end to end process handling that make us think a screen was required for this.

Ran the watchword, capture the details, and wherever the rules just don't match up, whatever energy will annotate the documents directly, for a very convenient way for somebody to review. Very interesting solution.

Customer is filing for an IP on this, because it is one of a kind in the industry, huge savings in ours, turnaround time, and, of course, reducing human dependency completely.

In fact, one of the biggest problems here was the container was incorrectly shipped that, with wrong information, wrong, you know, uh, yeah.

The wrong go, uh, goods shipped because of an oversight.

This would have been a huge, huge, problem because it's a huge loss for them, monetarily.

And, of course, SLAs were very important.

Um, Another very interesting one is in, and what we've done is, we've tried to take use cases in different industries. So you're going to labor of things across different areas.

This one is in, you know, in health care, It's a very common requirement to extract information from, know, your, your medical cards, and each, No provider.

And this is the US. And we're talking about equal lighter blue shoe grew, cross-check. Now, all of these have various different formats of their cards. So there are no consistency.

The amount of data that you need to extract is is consistent. But where the data right resides in that particular card is always changing.

Very big problem for this healthcare company, actually, is not healthcare's eyecare. So this I can accompany tag patient visits.

Then the scans will be taken and then they will be fed in manually by by no back Office stuff into their systems, but they wanted to automate this. So what we did was we were able to create a machine learning model to read all of these automatically and feed these into their systems.

They're now the team and only to cross verify this data across different systems. And validate and maybe corrective, if necessary.

So this helped tremendously impatient scheduling and all the other information on all the other processes actually, after patient scheduling, although no different processes that followed became absolutely seamless because this lookup was required.

Oh, another very interesting use case is or a gourmet catalog automation for a delivery network.

So this company was getting into the aggregator business for restaurants, and had a requirement to ingest all different restaurant menus.

And you can imagine this can be quite complex, because this is something, you know, if you look at any restaurant menu there, images, there's text. There's a lot of information that needs to be interpreted is not just data extraction. So, the item, the type of item, the, the additional options whether it's, you know, vegetarian and non vegetarian if it's done on the edge of this beef chicken, whatever. So all of these are different costs associated. They also have different menus and menu classifications.

That deserves the main course whatever, so all of this requires natural language processing, not just data extraction but you know interpreting that information and also organizing it in a manner that could be easily reviewed by somebody before it is ingested into the internal database.

So, very, very interesting use case again, and challenging one, because there were also images, which had to be reviewed, because there were certain requirements, whether the images should not have certain, no numbers on them in terms of dollar amounts and things. So, there were certain requirements that have to be met.

Um, another very interesting one is in medical query, this is a problem that requires a lot of human intervention.

So, every time there is no, an assessment done on a patient, and all the data is captured, this needs to be, you know, the medical coding for. This needs to be done by somebody who understands all the codes. And there are more than 70,000 codes based on which you need to interpret this.

You know, the treatment that has been done, and based on the type of rule, the size of wound therapy, the patient, or if it's, if he has some additional complication, All of those lead to different coach.

And very highly, complex, time, intensive, error prone Everything was automated by the AI powered RPA bots.

So, it will go in there, interpret the information, like a human, translated into an actual code.

And, these cords on to keep changing. You need to be up to date. So it will also keep learning new codes and and be able to relate to present.

And this was then, the ones, once it was coded properly, then it would go into the building. So, in the billing will lead to an external billing company and fit into that.

So, very interesting capability to learn on the fly and keep no adding different codes and learning conduit.

Another very interesting one actually, each one of these calls out a different area, maybe non traditional areas of automation.

So, we take these particularly because it gives them an IT, Oh, how automation can be achieved.

You know, most people think that automation only should be done in, in certain types of documents or certain types of PII related traditional areas that you can hear.

We have, these are some of these are very non traditional in the sense that the one we have right now is about the power company and meter readings, and this is something which is not very common, and we encountered this in Kuwait last year, actually There.

A company, a power company, had a problem that they were actually sending people in physically to take meter readings, many photographs and then interpreting them in and punching them back in into the sap System.

But, because of coded everything was in a lockdown and they couldn't do this anymore.

So, what we did for them wasn't ability to do meter reading through a phone app.

So, we built machine learning model for interpreting meters.

And we're able to get this deploy on to format which could leverage the own, itself, the infrastructure of the phone to do the extraction notion of what was required in the back end.

So this was the beauty in this was that now data distributed system every every phone became a point of entry for ingesting all this data for them.

Clean data I might add because they were extracted automatically. It will do the interpretation.

Any corrections could be done by the snow on food stuff. And then, once done, it was directly ingested into the sap system, became a very good end to end story for them.

Reduce conceivable saving, actually prompt.

From a problem, They created an opportunity. So the whole business model, It couldn't be changed based on this, that they didn't need as many people on the field anymore, because the customer could do this Now.

They didn't need an FTE to go a citizenship to go on the field. They could have the customer download this app on their phone, take a photograph, and that's all they needed.

So, lot of savings to just doing an automation like that.

This is another very interesting one.

This is an insurance, actually, Most of us think of boche as during lift and shift backend operations. We don't like to expose bought the business critical applications need to be implemented.

But this is one such use case where we actually had the bot operating in a front ending scenario.

Their critical business requirement wasn't addressed by the book, on the fly.

So then, when they don't, then somebody applies for a group.

Your group credit life insurance is something that is augmented to the insurance company, would like to sell Group Credit Lifecode for the loan.

And to be able to do that, they needed to do complex calculation. So it has to be done in real time when the loan is being requested.

Other words, it doesn't add much value because they were losing business because of the turnaround time. Initially somebody will request is. there would need 45 minutes to run a code generation to exo and other means, because of a complex premium balancing calculations, you know, cash flow, property margins, all of these calculations had to be done.

Especially because it was a group, no code.

So, because of this, the turnaround time was they would take maybe a day to get around and by the time, you would lose a lot of business. So, there was a viable business there.

So, what this company, I wanted to do was, something which is on the fly. That they could provide this, this code, on the fly.

And this was provided.

Who's done to watch So first, you were able to get the processing time down to less than NaN to the bot itself.

And that led to another, know, our use case for that matter.

Now, that bought itself to serve as this quote on the, then along with being requested to any partner that, you know, wasn't involved in this transaction.

So this led to a new entirely new product offering for them, the revenues of 60 million in the first year itself.

So this entire thing is running on our watch, and our application is running on our Bosch and it's very, very rare that you would see something that I bought this trusted to run an entire process of the commission critical business critical process.

Schraeder, that's a great point to ask a question which I have for you.

Do you have any video showing this bulk capabilities that you could share with the audience?

Absolutely. Let me do that. Let me, actually, there's one more case study.

This isn't wealth management.

I just covered this in a minute or less and then I'll jump into the video, but I can actually demonstrate what this is all about, right? So, you see an actual use case working?

Cool.

This is, this is about report generation in wealth management, where multiple reports were being requested and re query in are required to go out to external websites, You know, go to axial cheat. A little information.

This is a very typical vought requirement, and, of course, do a risk assessment to multiple different rules and regulatory violation, and, you know, penalties were were, you know, if it was not done properly, it would expose them to regulatory violation and penalty. So this had to be done, right?

And we were able to do this, and watch. the exceptions are less than 5%.

So, another very interesting use case, Very different use cases in different areas.

This brings me to exactly what you requested, so, let's talk about this demo. This demo is intelligent automation demo is going to touch touch upon all the different aspects of intelligent automation, that I talked about, that is for a loan origination process. Their documents. A new loan document comes in.

Firmware data is extracted using intelligent OCR and botch martial law codes download.

I will interpret the e-mail, understand the body of the e-mail, understand that this is a request, or a loan, and then download it, extract information from the document.

And then trigger a process flow that will go through multiple steps.

The first step being storing all this information Well first and being somebody reviewing the information extracted to make a checker scenario.

In a straight through processing case, this would have been bypassed. So, I'm going to talk a little bit through the demo as we're walking through it so, well.

Uh, I'll show you how this triggers. Like I said, this is, this is a control center, our product.

From there, we will schedule a bot and the bot will be periodically looking for new e-mails that come in and every new e-mail will be interpreted from the inbox.

The different kinds of e-mails will come in there, and we would interpret what.

Just like a human does, your interpreter can sort out those e-mails to ensure that only the ones it is processing for new norms would be the ones that first, you could have any, any amount of e-mail, and it could be a reply reply or, you know, multiple chain of e-mails, It doesn't matter.

It has the ability to interpret this like a human using machine learning to understand the intent, or what is being requested.

So it is sorting out the e-mails, depending on what is being requested into different, You know, folders and only, uh, looking for loan forms in the right knowledge has found the loan form, and it is going to now download this.

It interpreted this as a new loan request, is going to download this, open this up and start extracting from this in an intelligent way.

Just like a human, this is a personal loan application, where you're gonna look at, to applicants.

It is specifically highlighting the section on the left, to, to only extract from there, because that's where it's configured to do, is looking at all the different aspects, and you see the Red Highlights and the Blue Highlights merits finding anchors, and again, those anchor labels extracting the information to the right of that, which is the actual information that we need to process this loan.

What it's, what we're going to do now is log into the system into our control center and look for the first task. So, loan underwriting process, which has been preferred, has multiple.

So, we see you as a manager, I'll be able to see the entire audit trail on the process side briefly for a second that the first task was highlighted in in blue and that's on J Jacobs. Alex ...

is the person who would review this task and be able to do the extraction review.

So, here you will see that all the data, which was extracted is shown in a very seamless way, where the information contextually is highlighted the left. When you click on the left side, it highlights the right and vice versa.

So, the document itself is presented for somebody to read. Also. The accuracy of extraction is displayed to call out in red highlight only those problems.

You can also reassign this I'm sorry, we.

And Lilla Tumblr to back it up 11 to the point where we work.

So, um.

Give me a minute, and I want to also talked through this process a little bit.

So, we, uh, showed you how the highlights. Can we review left to right?

We will show you that briefly again, because it is moving the fastest wanted applause it for a second, to be able to give you a little flavor of that.

And before we move on to that, this process flow, with which it's going to quickly, I just wanted to pause here and show you that the first step in this application is a loan application data entry, which is what we were seeing a few seconds back.

And, once the person reviewed this, it couldn't go into a robotic legacy, loan origination system Update, where the data will be stored into a internal loan origination system By bought again. Then it will do other various checked, automated services. And then finally go down to a robotic appraisal task, where the robot would look for additional documents to process, then they come in, at a later date. So, that it will highlight this along the way.

So, every time it move ahead, it will be able to you will be able to see that, it is giving you the visibility on the process for a manager to see where it is in the process flow and how it is proceeding.

So, here you see in the task, all the data which has been extracted from the document is visible for you to see.

If you remember, this is the data that was extracted by the bot.

There you go.

And, this is a form that we were looking at earlier, OK, So, just wanted to make sure that you kind of a good glimpse of this, then we will complete the task, and as soon as we do that, it will move ahead to the next task.

It will trigger the Task four.

the robot, to take over 10 updates, so you can see now, it's highlighted the robot tasks, and the robot has kicked off, which is now going to be updating all this information into an internal loan origination system.

As well as a Legacy Loan Origination System, so both of these will be updated.

one of them is updated, knowledge will tailor another update tool.

A legacy system, which has just, you know, if it's 100 type of interface, now it moves ahead.

two, look for an appraisal report On the loan.

I'm the home loan. So the whole mass will be appraised, and then that's what comes in to the field.

And it is now downloading that extracting from that document and having somebody having the task assigned to somebody to review the appraisal details.

Of course, at all along the way, you have a full detailed audit trail of the fears and variables that were part of the process, when the task triggered and when the task ended. So, you know, who did what at any point in time, and we're the performer of this. This is the BPM particle flow.

And now you will see a more complex report, an appraisal report, which is going to be reviewed by somebody and once done.

They can either reassign or complete the task, and we'll go ahead and complete the task now.

And one's done.

It will then move further on where it will go through, you know, as a manager, again, I can see and review that it is moving along.

So, at any point, I have visibility on the status of lies and it is going to, you know, it executed some underwriting rules.

To me deciding whether it should be manual underwriting or it should be an auto underwriting, and it has now come to Michael, who's the underwriter, who will now review this. And we'll also be given some insights, machine learning insights on the fraud propensity of the loan and the confidence score of this. So, you see that 98.6% looks good, go ahead and we will approve this.

So, that's what happens, not as much further.

And as you can see, as it progresses, you have a detailed audit trail. As a manager, you can bring out reports, and you can also see, at any point where your data lives.

That concludes my demo.

But it gives you a good understanding of how bots can be added Innovate Intelligence, cognitive ability, RPA ML work together, and how VPN Keychain all of this together into an end to end process sloppier enterprise.

Fantastic shuttle. Quick question for mind. Do you have any domain specific solutions? Fantastic question, because that's exactly what my next slide is all about.

So, thank you, Ryan, leading me into that.

What a good solution is, something that are easy to build on, on this platform.

Because it is geared to build solutions, you can build invoice processing and then follow that up with the full full-blown accounts payable and then move to a P2P workflow completely.

You can also have a service desk automation yourself part of it in the loan origination demo.

It says that e-mails are being looked at and download and chartered and addressed could be in an insurance space. It could be lending.

It could be anywhere that you have e-mails coming in that need to be sorted, so that service desk automation, new customer onboarding, thank statement reconciliations.

We have a lot of use cases where we have solutions in insurance, Insurance brokerage, also, healthcare, eyecare, wealth management, all of these.

Know, so the solution lend it to does this platform lends itself very well to building solutions which are repeatable, and these are just some of the area and in many other areas.

We have built in many other areas too, but these are just a few that I can think of off the top of my head.

I would like to conclude with one last slide, and there are lot of questions that are asked of us that, you know, I want to start my journey, but it's risky. What if I don't succeed?

There are so many cases that I've heard of success, but so many equally we have not succeeded, So how do I really know that I'm on my on the right journey, how do I know I'm going to be successful? How do I take the risk out of no adopt, intelligent automation?

So that's where we have devised a quick start offering.

There, we have a guaranteed rollout overtrain bought in 3 to 4 weeks, So what do we do As part of this?

There are huge benefits. And the biggest one being taking that. Any risk? Taking? A strategic decision decision of this? Of this type, especially if you're making it for the first time and getting your key stakeholders to bind to it because there's a lot of convincing that we're going to start something new.

Uh, so this includes a dreamboat rollout for your solution in your domain.

That no obligation. And so there's a fixed price You would know upfront, but there's an obligation to purchase until you actually see the value and the Ottawa that you desire. So you, we develop it, we deployed, you run it.

You verify that it's actually giving you the tangible benefits that you are seeking and only when you are convinced, what do you actually, you know, have the commercial obligation an obligation.

But you would actually no get into the commercials to to continue using it.

So talk to the workshop goes into scoping implementation and rollout. It's a very tight deadline, 3 to 4 weeks, so you can know very quickly whether this works for you or not.

And the best part is there's no obligation.

So, taking the risk of April, making this kind of a decision is what we believe.

they'll have a lot of new, uh, new enterprises to adopt this. Especially those that are not very agile, they're not already invested very heavily in new technology, or enterprise wide automation.

So, Brian, did this address some of the things that you're looking at?

OK, I think Brian is on mute.

Uh, this format, I was just saying yes, it did, and I wanted to thank you very much for the past 45 minutes or so has been amassed a call soon, running people through this. We've now into the Q and A session, we've had some questions in already.

If you have any more questions you'd like to ask, sweat and water, you say, Please make use of this time. It's not very often you get a world-class leader in the space available for your personal questions.

So, please don't hesitate to put them in as we go through the Q&A.

Was there anything you'd like to add before we move into the Q&A?

No, I think we at least I've covered everything that I wanted to.

There's anything you have, specifically, before we went to the CAH Q&A I would be happy to address that.

Now I'm just very pleased that you managed to gain as much as you did during this time and I'm going to jump straight into the questions which we've had a tradeshow.

one of them is fairly obvious and almost hesitate to sort of ask this. But it's it makes sense from the point is how is the ... platform different from other automation solutions? What makes you stand out from the crowd so to speak.

Good question.

We get asked that a lot because of, uh, it is difficult to tell, you know, production Park, if you're not deep into the technology.

So, what we see is, there are other products out there that may have RPA and intelligent machine learning components, but may not have the third piece, which is a business process management.

There may be other products that have the BPM and RPA, but don't have the intelligent machine learning automation.

So, we see products that have maybe one, maybe two pieces, but we don't see those that have all three.

Some of them claim to have all three, but VPN, business process management has been around a long time.

It's not just human but orchestration as, uh, lot of folks, you know, local to that company, human, But orchestration is getting human and a bot to work together in a limited manner.

But having a full featured VPN requires no, the P P M N two O, that is that there is, uh, no, BPM in tool is something which is widely regarded as, uh, is it compliant?

If you have a engine, which is a business process engine, business process, management engine, which is BPM, and all 2.2 compliant.

Is the industry benchmark off having a fully featured BPM engine.

We don't see anybody else who has that. So each one of these is available separately, but as a whole. And in a manner that that covers everything seamlessly is, what we believe is missing in other products, that that also came to, have intelligent automation.

Fantastic. Thank you very much for that. I'm moving on into a question. It's very dear to my heart and I'm sure the same in every business person around the planet.

Essentially, and I'm sure it's a question that you are again, asked a lot but how do we measure the roi of an Intelligent Automation Project?

What should we be looking for?

Fantastic question, and, uh, This is, this is not simple.

Oh, 1 or 2 metric answer. So a lot of people start off immediately saying, hey, I want to measure it in terms of effort, savings, in terms of cost savings.

That is one of the direct things people look forward. It's not always as simple as that.

Many of the case studies I walk through did not have direct savings in terms of just monarchs are cost impact.

Some of them were much wider and more strategic than the one, which involves the container, you know, shipping of large containers. If they had one mistake in that container, they will huge. I mean, instead of making money on that, there were huge losses that we would encourage. So, getting it done correctly is very important, To the first time, right, is, is as important as, you know, unfortunately, you can, you can see that it's not done correctly. What, you're just going to end up doing more work, so, doing it right. Also, there are areas where you have SLAs. You want to do it quickly.

The answer is to be met.

You need to optimize the usage of your infrastructure, which could be not just your, you know, material components in your organization, but there could also be human elements.

So, all of these contribute, you know, using them in an optimal manner, giving you visibility, giving you audit ability, lot of your compliance requirements are good, amount of auditing, validity. All of this is important.

As otherwise, there's not just a direct cost implication.

Fantastic. And I'm sure there'll be a lot of follow up around that. And I'm sure that's a question that will will differ from case to case and from industry to industry.

My next question, I'm showing really relies, sorry.

Goes around the IDP capabilities of your platform. Can you share a few more details about that and for those that are not familiar with IT piece, expand on that a little bit more, please.

Sure. Thank you very much. Good question. Though. IDB is an area that is talked about, a lot nowadays because RPA is something we started with now we're moving, you know, getting getting more cognitive flavor to watch And that's very intelligent document extraction comes in because not only automation happens around document extraction.

We are a document centric you know, most businesses are document centric self, to be able to do that, a lot of products today.

No, Do different kind of pricing. one of the important aspects that you realize is, especially when an enterprise is a large volume, then you're paying per page extraction, It. It adds a lot of cost.

Pure automation, So we, we have multiple ways that we have addressed that. Per page.

You know, costing is something that you may eat into your ROI. So we gave us surveys license also, so those are some of the differences we read.

Being able to generate because eventually, if you do more ROI, you give more.

back to the customer, they will automate for.

You know, this is an area, process automation is never going to be done.

So the more you enable customers to automate, the more they will automate. It's a never-ending process.

So if you give them a means, You know, doing it in a more cost effective manner by not, you know, for example, charging per page and having a server that can do millions of documents.

Without that additional cost coming in, you can scale much more. So, intelligent document extraction is not about just document extraction.

It could be, you know, menu extraction, like we talked about could be meter reading.

So, machine learning models can be built for any kind of no no extraction from any image.

So, it works primarily through computer vision, machine learning on images.

Cheryl, thank you for that. And, it's always good to know the information around that.

And finally, one question that, I'm sure, again, you're all smart, and apologies if this is questions, which we, for you've had those responses, training and certification, G: Do you provide that? What, what, what is available from a negative perspective, please?

Absolutely. Good point again. Provide training, certification, all of this as part of your community portal, You will get trained, we also have classroom trainings.

Total customer, prefer one or the other, depending on how mature they are in this journey, and how comfortable they feel on our, on our platform.

And, of course, certification is, is something that you, you get certified after you go through the program and get trained.

So, certainly, we have that, and our focus is enabling customers, putting this in the hands of customer, and then, of course, partners to be able to, you know, do this automation on their own.

Fantastic.

Cheryl, thank you so much for taking the time to share your expertise and answer all questions during this presentation. Ladies, gentlemen, you can continue the conversation we show on social media. I have shared in the chat the link to address.

So that you can reach out to Shuttle afterwards. This presentation will be made available, on demand, post event. So please don't hesitate to download it and share it with your colleagues, and we look forward to seeing you within the next events in the series. And shuttle, thank you so much for joining us. It's very much appreciated, especially as you're sort of 4.5 hours ahead of where I am. And I'm in the UK. So, thank you very much, indeed, for your time and inputs.

Thank you, Ron.

Thank you. Ladies and gentlemen, if that ends the presentation for the day, please don't forget to download the handouts. Continue the conversation on social media.

Don't forget Sheryl's LinkedIn address is in the chat, and follow up afterwards with any questions which you may have which are not asked here.

Take care, Please keep safe and well.