On-Demand Webcast: Insurance Transformation Using Intelligent Automation
August 18, 2020
Speakers from EisnerAmper and UiPath discuss the impact that emerging technologies, like process automation, are having on the insurance community, and best practices for successful deployment and gaining firm-wide acceptance of these technologies.
Greg Fritsky: I've been in this space for five years, but I've been working with insurance companies for the better part of 25 years. Lots of different technologies, lots of different challenges, always seem to be focused on automating and deriving more value out of data. Very excited today to have Sathya Sethuraman join us from UiPath. He's a vertical industry leader within UiPath, a leader in the RPA space, partner of EisnerAmper’s, and very excited that he was able to join us today.
In addition to Sathya, I also have members of my team joining. Martin West, who is a solutions senior manager here at EisnerAmper. Marty helps lead a lot of projects around digital, and a number of different technologies and applications, and also Benjamin DiNapoli, senior solutions consultant. Also has actually been one of my original team members helping me build some of the early robotics process automation applications that we developed, and we continue to do today. With that said, I'd like to introduce Sathya.
Sathya Sethuraman:Thank you, Greg. Hey. Thanks for having me here. I am Sathya Sethuraman. I lead the global industry practice, insurance and industry practice for UiPath. Have been in the industry, I'm competing with Greg for years please in the industry. Having led a large digital transformation from advisory through implementation. Have been working with lot of insurance customers and prospects, and helping them to scale automation. Thank you.
Greg Fritsky:Thank you, Sathya. Today, again, you'll be eligible for CPA. Make sure you answer the polling questions. The goals today is really one, everyone to take away a few things. One is understand what RPA is. What are some of these key products that we're seeing as digital transformation takes hold. Everyone is discussing possible technology advancements at their companies. You may be evaluating something currently or just looking to use something better. Understanding how specifically within the insurance space, how does this help?
What are the use cases. There are many use cases, and we'll get into a few in a little bit. Most important thing I say with any of these technologies is understanding how to articulate and evangelize these technologies. We all have different understanding. We get excited, but sometimes we have a difficulty articulating it to other team members. How do we get programs going? How do we get people excited? How do we explain these technologies? I often think that that's the most important part.
You'll come away with an understanding, but how do you articulate it to others, and how could you advance your own programs? Today we're going to have a few things that we're going to talk about. We're going to start with drivers of digital transformation. It's pretty obvious to all of us that it is a remote and distributed workforce. Insurance is pretty much operating as business as usual, or unusual depending on how well equipment you were for the changes.
Having been in insurance space, I know that there are many different systems and processes. It takes a lot of people to make everything work within an insurer. Everything from claims management to policy management, investment accounting, even the core finance functions, operations functions. There is a lot of work that has to get done, and now most of it, if not all of it in some of you, it's being done remote and distributed. The other challenge is with the economics.
We're always being squeezed to try to do more with less, and people are not necessarily able to use the same level of workforce that they've had in the past. They're trying to do more, and not able to hire more. How do you find a way to automate and streamline? Probably something that's persistent for years is just that people are doing, and not necessarily thinking. What I mean by that is just think about the jobs that we have today.
How much of my role is where I'm actually producing a result versus analyzing the result? That's really at the core of a lot of these transformations, is trying to drive value and using people as intended, as originally intended, which is to have them think and decide and to make decisions. We'll talk a little bit also about RPA, but also how do you govern these projects? How do you make use case selections? How do you move these programs forward? Also, we're going to do a little deep dive into the insurance industry, the trends, the use cases.
Sathya will share his vision and insights having worked with customers, resolving a lot of their process challenges, and helping them find ways to drive more value within their organizations. We'll do a deep dive into specifically in the claims processing area. Drivers of digital transformation. Basically, I've seen for years going in, having done many workshops with clients, oftentimes we say business processes are complex.
People will say to me, "Our business processes are complex, and they haven't changed in many years. We ultimately want to be able to transform them, come up with some new future state that's streamlined, that saves money, that enhances value. I want better controls. I want better transparency." What we often find is that those business processes aren't necessarily complex. It's just that there are so many different parts to the process. There are so many different systems.
There's different disparate systems. There's different groups. We're built into silos and trying to bring that holistically across the enterprise. Trying to transformation that is often the real challenge. That's one particular operational challenge. The other is this drive to reduce cost, increase scale, doing more with less. The fact is that organizations were able to just keep adding bodies and solving the problem taking that approach. 20 years ago, we started to see a movement towards outsourcing and offshoring, looking for the same processes, but just done at less.
I've often heard people refer to it as your mess for less. Just passing the process on to another location and doing it cheaper, but still doing the same process. How do we automate that? How do we standardize that? In some cases, companies are looking to bring that back. They're trying to bring back some of those activities, particularly around control, about reporting, risk management. How do you start to get your handle on that, and how do you bring that back without incurring additional cost?
The last part is making work challenging. I think I've had a lot of cases where I've talked to clients that they employ people with advanced degrees, and CPAs, MBAs, and they're doing journal entries. They're posting entries. They're doing manual reconciliations. They're doing data entry, because that's the way our system is, and that's the way we're challenged. We're all using Excel, and we all know Excel, and we love Excel, but we do so much of our work in Excel.
Most of the work, when you evaluate it, is really what I would call data transformation, which is taking information from a source system, and then doing multiple changes to it until you get it to a level where it's ready to go back into another system. I'll see people spend five, six, seven steps running macros, pivot tables, making adjustments, and then just putting the information back into the system. That happens quite a bit, and that's our control, and Excel, it's a great tool, but it was meant to be more of a workaround for our systems, and it continues to be a big part of our operational challenges.
Martin West:Yeah. I would just add really quickly on this one, Greg, on talking about challenges, a lot of times, too, sometimes companies just want to throw a technology at something to fix an issue. More often than not, it could also be a data challenge, too, not just a technology itself. We'll get into it later, but it's really important to make sure you lay a good foundation of the process, have proper planning, and take a good look at your data before going along the journey.
Greg Fritsky:That's a good point, Marty. One other thing that companies are evaluating is if I'm going to take on these challenges, if I want to transform, how do I come up with the vision? Oftentimes people want to understand, "Should I be evaluating automation? Should I be looking at changing my processes? Should I enhance my controls? Where do I start?" Oftentimes I say, "This is a journey, and a journey requires an outcome." That's the most important part. What is the desired outcome? What are you trying to achieve?
Then ultimately making sure that that strategy aligns with what your C suite has set as the agenda. I don't know any organization right now, isn't looking for some way of streamlining costs and enabling their folks, given that they're working from home and being be levered to the technology better. Then it boils down to what is the right technology for you? What is the right application? What level of investment are we looking to make? Then how do we move the strategy forward?
These are important points to make because these programs, they require an initial evaluation, and then ultimately you need to make some sort of selection, some sort of decision, and then you need a roadmap to phase in these projects. Automation is something that you could start relatively quickly, but what's the end game, and how far are you going to take it, and how are you going to manage it? Then this is real important, is understanding the role of both IT and the business.
Are these really siloed groups? Oftentimes the business has the initiative, but the IT is setting the agenda. The business requirements don't always meet. That's a very important point, is how does the business determine what the requirements are, and how does IT decipher that information to build that plan out? It's very important to work together as part of that strategic vision.
Martin West:Yeah, I've seen that on both implementations, and also in some of the internal auto work we do, as well. When you have your business and you have your IT siloed, they don't talk well. They don't know what the other one is thinking. They're more familiar with what's important or the risks that they see in their own buckets. Unless you have them all at the table together, they don't understand the needs of each other, so it's really important to have everybody at the table at the same time, and get that good understanding across the board.
Greg Fritsky:In terms of transformation, and what we're going to be talking about today with robotics process automation, it's all about transformation. It's moving the dynamic from where you're spending very little time on analysis, and that's actually where you'll spend most of your time. The only way to get there is through automation, enhancing controls, and eliminating those manual activities. If you look at any particular process, if you white board a process, you'll find 70, 80, 90% of it is typically some level of just doing.
I execute these eight steps, and then I get the result, and I evaluate it, and I make a decision based on it. Most of what I do doesn't require judgment or any kind of research. It just is pulling the information from different systems to get the data so that I can make a decision. This is what we call process transformation.
Greg Fritsky:What I wanted to do now is with Sathya's help, paint the landscape here around RPA. I guess let's start in the beginning. What is it? An RPA is, I got into the space about five years ago. When someone had to explain it to me, the best way to understand it was it's a digital worker. It's somebody that can do just like you or I, and can process something end-to-end. Essentially, think of it this way. If you could define rules, and you said to me, "Greg, here's the ten steps I need you to follow to execute this process."
If I can lay it out in a structured way, and the data is accessible, and I have access, that basically lends itself very well to traditional RPA. Bots are being used not just to automate, but also can be used to transform data. As I was mentioning before, a lot of the activities within Excel that we perform today, pulling in data, extracts, and doing transformation can also be done with a bot. A lot of controls activities. It can also help with some of your workflow, sending emails, doing approvals, as well.
There are many, many activities. There's many opportunities to automate those different manual activities with a bot. Again, it's mostly suited for rules-based activities or structured data. However, we're seeing a change, and I'm sure Sathya can speak to some of this. What we're seeing is the technologies like UiPath have some embedded intelligence, and that's what this is all about. Moving beyond just process automation, and into the more cognitive machine learning capabilities.
Sathya Sethuraman:Yeah. Yes, Greg. When it comes to the digital transformation, as you rightly mentioned it is about the right tech for the desired outcome. Actually, the RPA as a platform, it's actually started for that to address the legacy landscape. It was predominantly at the startup. The journey, it was predominantly used as a patch of work for making data from point A to point B, but not really in terms of the data transformation, as you rightly mention.
Now, since many of the insurance executives have started looking at the success stories that were created by a few of the insurance who have delivered phenomenal success, and started expanding that workflow into end-to-end workflows, and we'll see that in our insurance later, started looking at the higher level of benefit. The executives have now started mandating their team to see that how we can embedded into the digital transformation.
As a part of digital transformation, absolutely. There is a need of our robots to work with human being, and also to work with the intelligent part of it. We just said AI, ML, and also other digital channels to work with cohesive is what we are definitely seeing in the market.
Greg Fritsky:How and where do we start with these types of technologies, and RPA in particular? Really evaluating that process, again, looking at what I would call the 80/20 rule. What are the activities that can be automated, if you will? It's an easy exercise of going through on a white board of what are the steps? If I'm performing a specific, like a bank reconciliation, and I'm extracting data from an external website, and I'm importing that information to Excel, and then I'm pulling down data from my ERP system, and then I'm performing an activity.
That's just a basic case, but the reality is looking at the inputs, how do you extract data? How do you process that? What is the output preparation? What type of master data is required, as well? What kind of setup? Are there internal controls involved? Are there administrative functions that you have to perform or execute. They're all very rules-based, recurring, and structured. Think of recurring journal entries. Think of reconciliations. Even exception processing, if it follows a rule.
The 20% are, think of the things that I have to do research. It requires judgment. It requires thought. It requires honestly to do the analytics. This is also where RPAs expanding into the more cognitive capabilities to provide even better insights, and the opportunities to make decisions even better and stronger. What I always say is technology is augmenting our intelligence, and not necessarily replacing it. This is an important point, because these tools really are there to help us enhance value and to drive better outcomes.
One of the things I often advocate, if you are looking to set up a program, think about the broad impacts. These technologies I've use cases where a specific business area will look to drive it, implement it, and then they're amazed when they see the results. The challenge always seems to be, "Now how do we govern it going forward? Who owns it? How do we manage it? How do we implement additional improvements?" I've been advocating this for a number of years now. This is really coming up with a center of excellence, center of expertise.
Whether you're a large organization or very small, it doesn't matter. You can have several individuals involved. You can have a handful. You can have one person assigned. At the end of the day, it's about having a combination of the business, or finance involved, helping define the requirements, having folks on the controls and audit side involved, because this is not just about enhancing processes. It's enhancing controls and transparency, and also having analysts.
They could be folks that are focused on the data and the outcomes, or it could be folks on the technology side looking to integrate applications better. No matter what you're doing, it's really having a multi-discipline. It's everybody's business to help automate. The best way is to have a group that's really focused on it.
Greg Fritsky:Okay. Manual effort sticks out in front, followed by complexity and lack of resources. It's pretty consistent with what I've seen in the past. Again, complexity of the business and why you're doing things, the way you're doing things is often a challenge. Lack of resources, it's more like are we using our resources to the best of our ability? Can we enhance? I think we find that we have very good people. We just have them doing a lot of manual work in an environment that continues to focus on cost cutting measures, and honestly data technology.
A lot of us have put in our systems, our ERPs. Now going on 20, in some cases 30 years or more. All those different ancillary systems, as well, that have to be integrated. This is interesting to me. We'll advance to the next slide. Actually, I'm going to go back one. One of the things that oftentimes folks, once I have a program in place, and now we're ready to get rolling, where do we start? That's often the next question. What are the use cases? Where should I evaluate?
Again, and we'll get into more specifics around insurance in a minute. If you're looking at a process, the best place to start is where are people spending that time? Where is that level of manual effort? It's usually front and center. More importantly is if you want to roll out a use case, and you want to do I would say an initial start, finding a use case where I wouldn't say it's difficult to implement. Something that can be stood up in a number of weeks as opposed to months is really where you want to start, getting that quick win.
Organizations really, they thrive when they see the return on investment is high. When the dollar is saved and starts to free up people from the day-to-day, they can spend even more time on these initiatives. That's always a logical, good place to start. Then taking a look at other opportunities to benefit from processes that might be a little bit more complex or require more investment. If the human element is very small, and it's a nice to have, then this could probably wait till a later phase.
First and foremost is trying to find something that's going to get you the quick win.
Ben DiNapoli:Yeah, Greg, I think that's a great point. For us, I know we spend a lot of time upfront coming up with use cases. Really, how can you see that ROI in the quickest manner possible, right? For example, we just implemented with one of our clients in operations a process that was touched by a few people multiple times through the week. It took them about three days, about eight hours in total a week. We were able to cut that down to six minutes they're now processing on a weekly basis.
Not only are they seeing the benefit organizationally from cost, time savings, freeing up some of their valuable resources to do other more valuable and strategic things, but from the employee perspective, they're really satisfied with, again, that's something that they no longer have to spend a quarter of their week doing. I think that's really important. The fact that they were able to see that upfront was huge from them from the top down.
Greg Fritsky:That's a good point.
Sathya Sethuraman:Yeah, actually, excellent point, Ben, because what we are also seeing is some of the places where there is a cost due to accuracy, also. Where we are talking about value versus volume that Greg was mentioning. There are high-volume tasks that we do in a very frequent manner which are low value. What happens is between those tasks of one minute or two minutes, what also happens is between shifting between one task to another task will take twice much as amount of time just to shift itself.
Actually the number of tasks that we handle in a day, if that can be automated, as Greg mentioned, to perform that seven or eight step other task, it might take one hour because we are consistently switching between one versus the other. When we see that we can cut it down, and also in addition to that, many times a lot of people are missing, a lot of insurers are missing was that is a cost to accuracy. An example I can give is an insurer who was doing automation in the subrogation process.
Just preparing the subrogation package, litigation package, and also following up on those subrogation cases and recoveries. What we saw was it is an everyday task for three of the operation people who are sitting and doing it every day repeatedly. The problem was when they send it out, they were having 60% of those process becoming erroneous in terms of data. They were coming back to them. The lawyers and also the assessors were actually field assessors, were losing time on coming back and forth.
These people were getting frustrated because they went through everything, all the seven systems they need to go through, and look at all the data, gather data, and send it back. Now it comes back to them 60% of the time remediate that process. It's an absolute value to roll out automation there in those sort of areas and look at them as enabling employees, and also reducing data, while also processing those cases much accurately so that your record data are more. Your litigation cases, the success rates are higher.
Greg Fritsky:Yep. Thank you, Sathya. I'm going to jump ahead to just tie down this whole conversation around transformation journey. Again, we've seen this evolution over time from implementing ERPs, process improvement. We've seen the rise of mobility and cloud solutions, and now we see robotics, machine learning. Again, we're moving towards augmented intelligence, the ability to use these tools to aid us, to help us make better decisions to have better outcomes.
This is all part of that transformation. Starting with something as simple as, "Hey, I want to automate these manual activities that we spend an awful lot of time with that we thought about outsourcing, but possibly we could just automate it, keep it in-house, and derive something from that of value," is a great way to look at it and start with those types of use cases. This is a journey, again.
Ben DiNapoli:Right, Greg, I think that's an important point to consider, just because the more we augment now, you could take it a little bit further than that original 80/20 we were talking about. Really the software like UiPath, for example, really does make it easy because not only can you start your automation journey there, but you can actually build AI on top of that within the same platform. Having that there is obviously a huge bonus as people look to expand it and improve their automation experience.
Greg Fritsky:Thanks, Ben.
Sathya Sethuraman:Also, this brings a very valuable point, actually, which we can actually pivot to where we are going to the next slides, also. What we have seen is traditionally, as I mentioned, the automation was seen as purely a task management effort where I can automate simple tasks. As the capability of the platforms grow, we embedded those into automation as the executives are also looking it to be embedded into the digital transformation, our option has grown. Not only that, what we have seen in the recent past after the COVID scenario is there are a couple of things that were happening in the industry already in 2019.
25% of the workforce have been already in the retirement mode in next three years. It is very difficult to replace those workforce in the insurance industry because people, the millennials, we were talking about millennials with men just before starting on this. Millennials want the cool jobs. They don't want to take routine, mundane jobs in insurance. Recently one of the insurance said that, "Okay, I am able to get people to do my claims operations, and also underwriting operations and sales operations when it comes to the summertime."
"The moment the colleges open, poof. Everyone is gone. My attrition rate goes between 35 to 45% during that." The reason being that having a relook at that, the future of workforce is what is going to be critical. Also, what I'm seeing is after the COVID scenario, while the insurers have started recouping themselves, and having adapted, the employees have started adopting to that new normal of the world, working remotely.
Still, most of the executives have still been saying that as the productivity has phenomenally come down. The productivity has come down up to 35%. This is the right time. Even those insurers who have not predominantly considered automation are doing a relook at automation to support those employees who are working remote as employee enablement, and they are embedding that into the future of workforce.
Greg Fritsky:Sorry. Go back to the original slide.
Sathya Sethuraman:Awesome. As we are looking at the problem statements, this definitely has changed from how many minutes, how many hours I am saving. How can I use automation to actually streamline my intake and underwriting process? Ben was actually mentioning about that. It's not just about the 80/20 rule, but how you can also stretch it with actually the artificial intelligence, machine learning, and also putting human in the loop? That has actually opened up the mind for many of the executives to relook at automation to actually embed it to these programs.
How I can do faster claim processing and do faster recoveries, or more accurate recoveries? Also, how do I wade through the challenges that I am facing in this current market? Also, in the insure tech are actually coming and jumping and disrupting the market. In those market disruption, how can I deliver superior agent and customer experience through our main channels? How do I take automation and use it as a lever in delivering those omni-channel experience?
Primarily the automations are now solving the problems of how can I sell more? How can I manage the risk better? When it comes to managing risk better, it is nothing better than finance and accounting operations. We have seen that finance and accounting operations have phenomenally adopted automation, and Greg, to your point, actually, what is happening in this specific finance and accounting space is the data transformation is becoming very, very highly important.
Again, accuracy is important, speed is important, and also taking away the employees with the mundane task, and shifting that time to actually analysis. Analysis in decision-making is always important, so making decision-making process so easy and seamless is what is critical, and managing the risk better is a critical point of automation. Then how do I operate it in a lower cost? As I was mentioning already, a certain set of population is retiring from the insurance industry workforce.
Now insurers are actually in a dilemma, whether I need to replenish those workforce and go through a churn. Even if the workforce is average salary in finance, and according what we have seen is between $58,000 to $75,000. Training, onboarding, off boarding, that cost actually doubles the cost of the workforce management. How do I manage the cost of the workforce? By adding robots to my footprint, and also providing digital assistant to my employees.
These are some of the broader and higher level benefits that insurers are seeing. As we move to the next slide. We'll also see that how across the landscape the automation is being used.
Sathya Sethuraman:Yeah, it is definitely very, very insightful. I would say it also resonates with the number of people who are shifting from the, who started with the finance, accounting, and reconciliation are, and how they are moving towards some of the core processes. This clearly shows that shift that has been happening. Again, this is not a one and done scenario. There are some insurers who are purely looking at the growth, are looking at underwriting and policy administration area, while there are insurers who are actually well-established and have done some sort of transformation are seeing that even with the modern systems such as Gateway and that sort of systems.
There were a lot of gaps in processing the claims, and that is a phenomenal way to deliver automation in those claims area. Greg, you want to chime in here?
Greg Fritsky:Yes, I also agree. I think the reconciliations piece continues to be problematic for folks. Again, a lot of that goes back to the data transformation piece. Going through Excel, making sure that the data is validated before making decisions. It requires a lot of time. It's very consuming, and that doesn't surprise me that we still see almost 50% responding that reconciliation is one of the most tedious tasks.
Sathya Sethuraman:Yep, yep. Also we are seeing, as I was mentioning about some of the mature insurers have started deploying robots in the areas of license management, agency compensation management, and reconciliation. Creating agents dashboard with respect to managing their future pipeline, customer services, and also reconciling their commission statement. In the new business and code scenario again, the insurers have started looking at intake of ACORD forms and processing of the ACORD forms.
They were starting there and expanding into email classification, whether it is new business or customer service or claims on the upstream. Also on the downstream, aggregating data from multiple sources to create a better and accurate core processing, and delivering policies faster. Policy quote to insurance ratio, and quote to insurance time need to be better. How do I operate agile so that I can sell more? Again going back to the selling more point.
Also when it comes to customer service, there are insurers who are deployed in doing simple customer services that includes customer address change management, and also doing some of the endorsement, non-monetary, and starting with non-monetary endorsement, but expanding into adding vehicles, deleting vehicles, split of the accounts, so those sort of scenarios. Claims is obvious. As we've seen, claims is number two in the poll. We have seen that claims processing, first notice of loss handling through payment processing.
One of the insurers has done million payments across the globe. This is a global insurer, million payments across the globe in their shared services claims operations. Also we have seen, I already mentioned about litigation and subrogation, and of insurers have seen phenomenal amount of recoveries from the subrogation, and also better win ratio in the litigation packages. Obviously there are other operation process that includes investment management, finance, and accounting, HR processes where customers have been implementing the automation journey.
As we move to the next step, we'll also see some of the business imperatives. Addressing some of the current challenges very specific to claim processes. Again, as I mentioned, predominantly it started with saving few minutes and few hours. Now they are looking at how do I actually resolve customer's moment of truth when the evolving customer needs are looking at Amazon for a service experience or claims experience? How can I sell them better? Also, how can I solve your customer problem?
Many customers who want to have multi-channel experience, how they can actually send an email, and also go back. Immediately a robot takes care of that email request, and logs the claims request, and post it back into their portal, sends an email to the customer. Now the customer can just go to the portal and check the status, because we know that the status checking itself is a phenomenal... I would say putting pressure on their contact center. For every claims that is submitted, actually insurers get seven to nine calls just to run the status check and follow-up.
Now by pushing that left to customer to do the self-service, the robots and automation can help you to deflect those calls into self-service so that the contact center representatives are much more geared towards serving the customers with the more complex questions and fulfillment journey. We can move to the next step to see some of them in action. These are some of the case studies that we have seen. Customers have moved from the old method of manual task where the customer need to wait to handle those, or get answers for the questions, and also settle the claim.
Now fast-tracking that 30, 40% improvement in operational efficiency, and also lower customer attrition ratio, and higher customer satisfaction ratios. Now we will move out of the possible demo that we will show where the insurer can use the virtual assistant and the remote assistant seamlessly. If you could see that HR part that is posted as a part of an insurer's website, can seamlessly talk the remote assistant, which is an unattended robot in the background. You could see that in the left.
The customer, the claimant that actually comes and asks, "I want to submit a death claim." Because mostly we know that the people who are all about 40 years age, they are very comfortable with the chat bots, because that also seamlessly connects them to the web chats. Here the chat bot can handle simple queries and simple transactions where it can start interacting with the customer with a simple question, asking them, "Hey." Similar to what a web chat will ask, right?
"Hey, what is the name of the customer? What is the policy number?" Capturing those information. Now once that is done, the robot will start triggering all the backend processing to go to policy administration in the background, and start kicking off that process. We also know that since the insurers have acquired systems in the background, they may have one or multiple administration systems, one or multiple claims systems totally working differently.
Now, though we have assimilated with the CRM, but you can now visualize that this is a policy administration system where the robots go and search for the policy, and takes customer information. Hold on to it, send a two-factor authentication to the claimant through a different channel that can be in a text or through email. Then once the customer provides that, the chat bot providing more information. Then the chat bot now knows that what other information I need to capture for initiating the claim?
It asks the other questions. "Hey, what is the place of death, date of death?" Once that captures all this claims-related information, it can also ask for uploading the documents. There are some minimum required documents that are required. That can include death certificate, ID information that are required. It can again comply with, if it is a medical claim, it can add the HIPAA compliance. In other places, it can fully comply with, I know GDPR and other regulations because it is fully authenticated and two-factor authenticated.
Those information will be stored by the robot in a secure manner in the backend. Once that is done, the robot takes all this information, does OCR, an optical character recognition, and actually pulls the information. Now it has got policy information that is gathered from the policy admin system, the claims information that is gathered from customer, and also additional information gathered from the documents that are acquired from the claimant.
Now it goes to your claims system, inputs all this information, preprints the information, generates a claims ID, and then communicate back to the claimant about the claims acknowledgement. Not only that, here it can also generate a claims ID and send it to their preferred channel. That can include text or email, through which they can communicate. Then from that link, the customer can do self-service, either coming back to the chat bot, checking for the status, or they can go back to the portal that the robot provides as a link and start checking the status.
Here, you can look at it. If it is a simple process of non-early claim, that is where the policy has been life for more than five years, and if it is less than $50,000. Then with the simple documentation, the robot can just send all this information to a claims adjuster, and they can have an eyeball on those claims, and then take quicker action. If it is a much complex claim which requires more information, for example the claims that are in the non-disability clause, then you will likely require more information.
That might be a physician statement, an actuary statement, and also your employer statement based on how many days you are on sick leave. Those are additional information. Even there, yeah, another robot can actually base on the rules, it can actually trigger subsequent requirement to gather those subsequent requirements. If you see that, that is the robotic capabilities are totally expanding from a simple moving data from point A to point B to actually starting with the multi-channel environment of chat bot, virtual assistant, to email, text to a contact center in the background, and also communicating with the claims adjusters to actually go for the downstream processes.
That is what even Ben was mentioning. Ben was mentioning that, hey, you can expand that 80% into 90% as you are looking at these high-value processes and trigger downstream. Ben, you want to chime in here?
Ben DiNapoli:Yeah, I think from our perspective, you hit it on the head, Sathya. I think as these technologies continue to be enhanced, I think we're just seeing a lot more capabilities in that area. I've been working with the technology now for a few years, and when we first started developing these solutions, again, there were limitations to the technology. I think specific to insurance, you saw right there. There was an intelligence component to it because the chat bot was driving anything going on the backend.
It did make some adjustments, and did finish off the process itself. I think that just goes to show you, and why we wanted to show at demo is really how can you start with a base automation to a pretty common, I don't want to say simple problem, but it's a process that a lot of us insurers will do. Again, it took it kind of end-to-end, so I thought that was a great demo for you to show there. All right, we're going to go right into some Q&A here. I know we have about ten minutes.
Please do send any question from the audience. If we can't get to those live, we will certainly follow up with you after and answer any questions. You also have our contact information on the screen. For the first question, Sathya, why don't you kick us off here? We talked a lot about where to start and how to implement these solutions. In your opinion, what is the best practice approach to implementing these types of automation solutions? Does it start with tone at the top, and again, that working relationship between business and IT? How would you describe that?
Sathya Sethuraman:We can approach it both ways. One is top-down and bottom-up. In both places, discovery is very important. That's right. When we talk about the top-down, it is using of process mining, where the IT actually takes the end-to-end process that are done in, say, gateway system. ERP systems like SAP, and analyze those process, and see that which are all the areas where the process have bottlenecks or choke hold? Then rolling out automations in that place.
This is more like a top-down. In the meantime, many times the businesses have starting looking at bottom-up and started rolling out, what we call it is Automation Hub, which is a platform that helps to manage those pipeline, and also go to citizen-led development where the people or the citizens are actively contributing ideas, and also using what we call it as task catcher, where the citizens are just recording whatever they are doing, the micro-tasks, and then uploading those ideas.
This is both ways. One is from IT standpoint where it is driven top-down. Our center of excellence, what we call it as center of excellence, which top-down, and the bottom-up where the business enables those back logs and pass it on to the IT to start building those automations.
Ben DiNapoli:That's a great segue into what I wanted to touch on next. Just one of the follow up questions was again, I know we talked a lot about starting with that center of excellence, and also picking the right use case. Greg, in your opinion, what's the first step that you really need to hit a home run with in order to get these bots one, going from a discussion perspective, and then two, moving along the process into production?
Greg Fritsky:Yeah, absolutely making sure you pick the use case that's going to give you maximum ROI right out of the gate. People are believers when they see what can be achieved. Our recent use case you and I worked on where 160 hours can be translated to minutes. That incredible, and those are the kind of results you can see. It's amazing what can be achieved, so absolutely picking one right from this get-go and hitting it out of the park, but also to Sathya's point, making sure you look beyond the success of that initial bot or two and building a program without the guardrails and the milestones and the outcomes to find, these things become a challenge.
Early success will just slow down. This is really a process improvement play, and all areas are in scope. Absolutely focusing on where the pain is and where people are spending the time, or wasting time, I would say, and you can derive much more value from their efforts.
Ben DiNapoli:Great. Absolutely. Again, our last use case, that was a real great one. Based off what I was saying before, again, the benefits are clear. From my perspective, seeing from the employee side because we all know sometimes that can be a sensitive topic of we're automating some of our daily day-to-day tasks. Again, in that case, they were actually a lot happier after the fact and noticed the savings. They were saying, "No more working till 8:00, 9:00 every day." Definitely great from a use case perspective.
This one could be for all of you guys, really, just in terms of what you've seen. Have you guys noticed or has there been any pushback in terms of sharing private information or PII over a chat bot or another automated solution? Sathya, if you want to kick that one off, and then let the rest of the team share some insights, as well.
Sathya Sethuraman:Yeah, sure, sure. Definitely. We have seen even the health insurers - PAA is one part of even, I would go much beyond. There are PCA requirements and HIPAA, much more stringent requirements. Customers are willing to do that, and are willing to share information as long as we provide or the insurers provide secure way of communicating with a customer. That is where we were talking about the two-factor authentication, number one. Second thing, reinforcing those security, and building those trust with the customer is very, very critical.
We have even seen, it's not just us, but also the big four like Mackenzie, EY. They have gone to customers and asked these questions. The customers absolutely trust chat bots much more than even the mobile channel because they have been providing this to web chats. They treat it as web chat. Again, the most important thing is to reinforce that security to the customer, and also build that trust with the customer. These are very, very critical factor for the customers to actually share this information.
Ben DiNapoli:Yeah. I'll just concur with that. I think again, it gets back to governance and controls. It's like any IT initiative. There's always going to be concerns around security and data and information. It needs to align with your company's policies and procedures. Certainly putting a layer of governance around that, and that's why creating that center of expertise with a layer of governance is very important, and can be a very powerful tool.
I would say that there are definitely safeguards you can put in place, and absolutely a priority.
Greg Fritsky:Yeah, and just to wrap up on this one, again, I think something that's sometimes overlooked from an automation perspective is streamlining these processes and taking the human component out of it actually does reduce the risk associated with those processes. Whether that's on the control side, it's tightening up controls, getting rid of some exceptions, or just in terms of processing. I think that was a great point to wrap up on.
Ben DiNapoli:I think we're at the top of the hour. I want to thank everybody. I want to thank our guest, Sathya. I really appreciate it. Great partner, and definitely an industry leader, and I thank you for your time today, and I thank all of you for your times. Please stay safe and enjoy the rest of the summer.