Automation for insurance carriers and brokers
September 16, 2021, 12:00 PM EDT
September 16, 2021, 12:00 PM EDT
Constantly extracting and entering policy information into your systems, moving data between software platforms that don't communicate, and other heavy administrative work - there are certain tasks that are very important, but are expensive and no one likes to do.
These tasks are repetitive and you can better spend your human resources - talking to customers for instance. How can we fix this? Automation is the answer.
By using a digital workforce, companies can save money, get back precious time and refocus employees on customers, not administrative work. Become more competitive in a very competitive market.
Our panel of insurance leaders and automation experts will demystify automation and help you understand how to implement a digital worker in your business and add directly to your bottom line.
Automation for insurance carriers and brokers - Robocorp x Q
Speaker: Peter Steube 00:00
Thank you for joining us. My name is Peter Steube speaking on behalf of Robocorp. You are listening to our shared webinar with Quandri as well as Brock Longworth of Cornerstone Insurance Services. And we're going to be kind of going over all the ins and outs of RPA and its applications for the insurance industry today. So again, my name is Peter Steube. I work with Robocorp. We are a company that is enabling the next generation of automation. I'm proud to be responsible for our global partnerships and I'm honored to share the stage with one of our top partners. They are early adopters also innovative entrepreneurs and not to mention they're also related as well. So, co-founders of Quandri, Jamieson and Jackson Fregeau. They have graciously enlisted the presence of their customer. He deserves substantial amount of credit for his own vision to embrace automation as a valuable tool in battling industry wide challenges and competition that is Brock Longworth, CEO of Cornerstone Insurance Services.
So, over the course of today's webinar, we will be touching upon a broad range of topics which are designed to have something for everybody whether you're an automation veteran, a new entrant or insurance practitioner and everybody in between. So, we're going to start out with a brief introduction to Robocorp, our technology and how it's helping Quandri deliver the future of automation. Then also, we're going to hand things off to the brothers themselves for a business overview, an introduction to what exactly is RPA, the concept of digital workers and how that's augmenting workforces as well as some of the results they've been able to achieve for Brock and other insurance brokerages and providers. Our arguably most exciting portion is a panel Q&A with Brock taking center stage to discuss his experiences pursuing automation with Quandri and some guidance for those who wish to do the same. Then we will wrap with audience submitted questions so please feel free to place them in the chat or the Q&A section of the webinar here and we will hopefully answer all your questions at the end. And definitely, we hope that the people here with us now stay on the entire time. But keep in mind, we will have a replay published soon afterwards. So, for those of you who choose wisely to work with Quandri, you may never interact with Robocorp directly and for us, that's okay. We're thrilled to be the technology platform that enables minds like Jackson and Jamieson to bring high ROI automated processes to life. For those of you who are curious just who we are, we're a startup who's just beginning to transform the robotic process automation industry but we're also very much the real deal.
In over the past two years, we've taken on in excess of 30 million in US dollars funding from some of venture capitals best and brightest. We've also on boarded 1000s of developers around the globe and touched organizations of all shapes and sizes. I'm excited to share today that these metrics are compounding in multiples not just percentage points. So, we have a great headway ahead of us and plenty of opportunities to continue to support the efforts of Quandri.
So, one might ask, why are you so special? Why have you raised so much money and achieve such great things in such a short amount of time. For those of you who have worked with automation before, it is very likely that this slide will resonate with you. For those of you who are maybe learning what RPA is and just also recognizing that it has been the fastest growing sector within technology for quite some time now and you want to be involved, you might not realize that behind those great headlines which are still very much true and what it's able to achieve exists some lesser known, we'll call them dirty truths about automation in practice. To date, RPA has been a market dominated by what are called proprietary brittle or expensive tools to bring automations to life and they typically have a vision that every user can build, maintain and run bots. It's more or less been a luxury afforded only by large organizations and even for them, there's been deployment struggles, sustainability issues and ultimately false promises of fast and carefree ROI. For us, Robocorp’s approach is different.
We don't disagree that everyone should have access to automation or better yet be impacted by it. It's more or less that we just think especially as automated processes become more complex and unwieldy, the development should be left to professionals like Quandri. So well, we don't necessarily disagree entirely with the vision of “citizen developers or users creating their own automations.” We just think that at a certain point of mission criticality or complexity, it's best to leave the building and maintenance in the hands of experts. And well, line of business users can focus on ways to use their newly discovered free time as they wave goodbye to administrative, repetitive and mundane tasks. Altogether, for us, the antidote to the challenge is highlighted on the previous slide we believe lies in what's programmable, open source flexible and more or less a modern approach that is more sustainable, easily scaled and fairly priced technology. So, when it comes to Quandri’s projects, it's both important and valuable to know that they aren't alone. We have invested significantly to build an “enterprise grade platform” that is accessible by all. Our infrastructure is fast to set up, elegant and it's intuitive for developers and includes modern architecture features like CICD pipelines, GitHub integrations and it's built on serverless technology for those of you who are aware of what that might be and the best part is we have our entire team, venture backing and expertise to keep it up and running smoothly. There is also built in security that is proven for some of the world's most discerning organizations. There are also dashboards and one place to monitor all the bot operations effectively including secure access control and user management and our community is incredible. Open source is a mentality that extends beyond just our business model. We have a legion of active members providing help when needed and problem solving and innovating together.
And lastly, when it comes to support for business models and processes and all that you may want to automate, we have an extreme number of libraries and also in alignment with the coding language python which is very ubiquitous and makes it very easy for expert developers like Jamieson and his team to bring anything to market.
So, with that said, next, we're going to hand things off to an introduction of Jamieson and Jackson. Jackson Jamison, thanks for joining us today and we appreciate everybody sitting through the Robocorp portion of the intro but we'd love to kick things off with your story. Where did Quandri originate and more importantly, did you start with another RPA tool or are you fresh on Robocorp? Why don't we get right into it there as well?
Speaker: Jackson Fregeau 07:52
Sure, thanks a lot, Peter. So, as Peter related to, we are brothers. So, I'm Jackson and the other co-founder here is Jamieson. And when Quandri started, Jamieson and I have wanted to go into business together for a long time so we've been looking at different business ideas for a while, kind of waiting for that right opportunity. And where Quandri originated, I was at another company and it came out of a real-world problem where we had really high volume, repetitive operations. So, we were a sales and marketing company and we had 1000s of emails that came in every day and we had people that needed to categorize and respond to those emails and they had to bucket them and we had a pretty large operations team dealing with that. And we knew that there had to be a better way to do it but we couldn't really figure out how to do it properly. So, after a lot of research and looking around, we landed on this bot building technology but actually with the different tools. Some of you may know them, UiPath which is an older version of RPA software and which Robocorp is now displacing tools like that. But what UiPath allows you to do is to build these bots with drag and drop software but it's a little bit clunky and it has really high upfront costs. But we built a few bots with UiPath and even there, we're still blown away with the results. And then Jamieson and I started talking about it and we came to the conclusion that there was a fairly large underserved need in the SMB market for this kind of technology that typically been reserved for large enterprise organizations.
So, Jamieson and I fairly quickly decided to go into business together and fairly quickly after that, we found Robocorp which really fit the business that we want to build as there weren't high upfront licensing costs. It was usage pricing and you can build all of the bots in Python. So, Jamieson's a developer, we want to work with developers and allows developers to do what they do best which is code. So that was about two years ago at this point. Since then, we've really focused on the insurance broker market. So, we've been with brokers all across North America at this point and all of this is based on Robocorp’s technology. So, we've been growing quickly focusing and helping insurance brokers and that's why we're here today because we want to be able to get this technology into the hands of brokers and allow them to continue to benefit from this kind of technology.
Speaker: Peter Steube 10:18
Excellent! Thanks, so much for that Jackson. As we kind of dive deeper here into what is robotic process automation, how would you guys define that and what problems do you think it's solving and more or less, why would somebody pursue automation?
Speaker: Jamieson Fregeau 10:38
Yeah, so while we're in the business of building these bots, really, at the end of the day, it comes down to the people. So, as Jackson said, we were starting and looking at this industry and providing automation in many different industries. And what we found when we were looking across industries is that people were spending an enormous amount of time on these high volume, repetitive processes. And Quandri is really born out of is the idea of taking the best of people and the best of computers and putting that together and you're able to achieve something you wouldn't be able to achieve without either of those. And that really is, in our minds, what automation and what robotic process automation is in the most tangible form of taking those repetitive processes that people really shouldn't be working on and putting computers to work on that so that people can work on the things that they should be like talking to customers, collaborating with other people, problem solving, all these really incredible skills that people are great at but in so many industries and so many jobs, people are spending their time bogged down on these high volume processes.
So, for us, robotic process automation allows people to essentially get that work done but enable them to focus on the skills that they're great at.
Speaker: Jackson Fregeau 12:04
And one thing we like to say in that same line is really taking the robot out of the person and giving that work back to the robot. You don't want that person doing the robot work, let the robot do that work. And ultimately, like what these digital workers are, they're automations that are mimicking what a person is doing. So, they're logging in with the username and password, they're clicking over here, they're typing over there. They're running that same process but they're doing it hundreds of times faster for 50% or less of the cost and they don't make mistakes or take vacation. It allows people to really focus on what they do best which in a lot of brokers cases is talking to customers.
Speaker: Peter Steube 12:50
Excellent! Thanks, guys for that additional color. And of course, let's have actual data do the talking here. So maybe you guys can walk us through some of what you've been able to accomplish for the insurance industry and some of the benefits that brokers are seeing. I'm sure Brock will also chime in as part of the panel here very momentarily but maybe you can walk us through especially the case study that we're showing here on the screen which is available on Robocorp’s website as well. Kind of talk about some of the results that you've been able to drive.
Speaker: Jamieson Fregeau 13:22
Yeah, and really initially, I think there's important reasons why we ended up in the insurance industry. So, as Jackson alluded to, we started doing this across many different industries and we're doing that to see where these digital workers have produced the most value. So, as we were doing that, that led us pretty quickly into the insurance industry because there are a large number of high volume and repetitive processes. They're super important to do but shouldn't necessarily be done by people.
And so, as we're looking at the insurance industry, we think this is because that there are just a lot of different technology platforms that brokers and insurance companies use but there isn't necessarily a way to connect these platforms. Now, a lot of these companies are saying they're coming out with an API or a way to connect the platforms together. But what we're trying to do is provide a solution to that now where essentially as opposed to waiting for five or 10 years for these companies to come out with something to connect their platforms, we can build a digital worker to essentially move data in between these platforms and alleviate these high volume and repetitive processes. And really a great example of that is this case study we're looking at now. So, this is a case study with Cornerstone that we'll dive into a little bit here and essentially, where they had an employee spending up to six hours a day downloading and renaming policies. Essentially, we were able to build a digital world to come in and completely automate this process so that that employee can move on to higher volume processes and work.
So, on the staff side, we've now been able to cut the cost on this process by 55% while saving them 120 hours every single month. This process is completed overnight so that when they come up to work in the morning, it's completely finished and it is finished 88% faster overnight. What this enables the employee to do is to go and focus on those more important processes like speaking with customers and generating new revenue for the business.
Speaker: Peter Steube 15:51
Excellent, thanks so much. And our final slide here before the panel Q&A was to discuss the relationship between Robocop and Quandri. I was kind of giving my bias pitch of Robocorp here for the first five minutes of the call but we'd love to hear from you guys at Quandri, why Robocorp? Why is this where you're building your business or why are you moving forward in using our technology for the folks on the line that might be interested in that relationship?
Speaker: Jamieson Fregeau 16:27
I really do think you hit the nail on the head with what you're saying. As Jackson said, we started out with another RPA platform. We started out with UiPath and as we were diving into that, the entire time we were working with them, we were thinking about why we couldn't just do this on our own in programming languages in Python.
So, as we were looking at that, and looking for ways to do it on our own, where we aren't restricted by the actual platform they make you build in and they are extremely expensive upfront cost, we came across Robocorp. And essentially, there's really a few things that make Robocorp stand out and one is just that it's incredibly flexible. We can bring in any technology we need to get the job done to build a digital worker and it's imported by Robocorp because Robocorp supports the most popular programming language, Python. And the costs are extremely reasonable because we only pay for what we use and when we upload our digital workers to Robocorp, we don't have to pay any strict licensing fees and it's all baked into the digital worker.
Speaker: Jackson Fregeau 17:44
And just to add one thing onto what you're saying together, ultimately, what that allows us to do is that allows us to pass cost savings on to our customers. So instead of us making our customer pay a high upfront cost, we can have a small upfront development fee and then we can run that on a monthly subscription sometimes based on strictly what they use. So, it allows us to provide a much more flexible pricing model to our customer and ultimately, that's why we really like Robocorp.
Speaker: Peter Steube 18:13
Great! Thanks, so much for that, guys. Now, let's move on to our main attraction here. Brock, thank you very much for joining and I know that you know Jackson and Jamieson well. We've had a chance to get to know each other just a little bit but very excited to hear your insights. And I'm going to hand things off mostly to the guys at Quandri and yourself to kind of take the conversation from here but wanted to say thanks for joining the line.
Speaker: Brock Longworth 18:38
Thanks for having me.
Speaker: Jackson Fregeau 18:40
And thanks a lot for the introduction Peter and going through all of that. It's really helpful for us to see that it's everyone's smiling, I'm sure for everybody on the line too. And now to Brock maybe to kick things off Brock, why don't you just give us a brief description of Cornerstone your business, how you fit into the insurance market?
Speaker: Brock Longworth 19:00
Sure, yeah. So, we are an insurance broker. So essentially, our function is to attract and interact with customers and then we will present those customers to the insurance marketplace, securing insurance products that meet their specific needs and then we will ultimately attempt to make that sale to the customer and with that insurance product. So, we source the products from the entire marketplace and then ultimately make a sale. So, we are a sales organization at the end of the day and we're very much driven on relationships with the customers that we have and relationships with the customers that we're trying to attract as well. And there's one little differentiator with us like we are a traditional bricks and mortar broker like we do have five off office locations where people walk in the front door. But we also have another brand called Swift Digital Insurance which is our online broker brand operating in the four western provinces of Canada.
So, because of Swift we've always been or since Swift launched in 2018, we've been pretty focused on kind of technology and what's next and how can we use technology to enable the insurance process as best as possible. I think that is ultimately how we ended up with the guys at Qunadri and how our relationship tickets serve with them.
Speaker: Jackson Fregeau 20:13
Thanks, Brock. And I guess ultimately like what was that? You mentioned having that digital brokerage swift kind of pushed in that direction. But with the digital workers or the bots specifically, what really pushed you to look for that kind of solution or what was the pain point that you were feeling in your brokerage to push you in that direction?
Speaker: Brock Longworth 20:32
Yeah, I think like a lot of insurance brokers like we are suffering from an inability to attract staff. So, we're having a really hard time finding people to come in and work actually in those high processing manual jobs. Like there are 1000s of repetitive clicks and types and touches in some cases of our business. So, it's hard to find someone that wants to sit in the basement and just type all day. So ultimately, it was just an inability to staff those positions that was the driver for us after engaging with you guys in the process. So, we did realize there was going to be cost savings, accuracy, reliability. There's a whole myriad of reasons why it made sense for us once we actually dove into it further. But the initial driver for us actually wasn't inability to staff those positions.
Speaker: Jackson Fregeau 21:19
Hmm. Interesting. And then I think what we're going to do here is insert a little bit of a poll and see from the audience who might be experiencing some of those pain points themselves or who is having that inability to staff or a difficult time running those processes internally. Great music Peter.
Speaker: Peter Steube 21:42
I think we can leave the poll open. We have respondents yeah. 12 out of 12 have said yes, they agree Longworth (21:54) so let me just cut it off there and recognize that a lot of people are experiencing the same thing.
Speaker: Jackson Fregeau 22:04
So that was 12 out of 12 were experiencing the same thing and that's what we've noticed with a lot of our conversations as well with brokers is that a lot of these businesses, they are similar and you see a lot of these problems across the industry. And that's ultimately why we're building Qunadri is to solve those pain points. But maybe Brock, it'd be good for you to kind of describe to the audience what bots that you have built to solve some of these problems.
Speaker: Brock Longworth 22:39
Yeah, for sure. So, the first thing that we looked at was, basically, as an insurance broker, we interact with a dozen insurance companies. Those dozen insurance companies work with hundreds of brokers. Those hundreds of brokers use half a dozen different broker management systems. So, there's multiple companies using a ton of different kinds of software and we're all trying to accomplish a similar task. But as Jay said earlier, there's not a lot of integration between those different companies and software’s. I've been in the industry 20 years and it's kind of been the holy grail of insurance companies and insurance brokers that we're going to find this seamless integration where this one software system we all play nice in and it hasn't happened and I'm not holding my breath it's going to happen before I retire. So, I do think that there needed to be something that could kind of bridge that gap for us. And another issue we have too is that just because I want to present myself to my customer this way doesn't mean broker B, C, and D want to present themselves in that way too. So, if ever the industry does reach that situation where we have that harmonious system that we're all using, there's a risk that we don't have the ability to differentiate ourselves as well.
So that's kind of I guess the background on why we wanted to automate our e-doc process which is the first process that we automated. So basically, our insurance companies will download our client’s policies to us. They're downloaded into formats whether it's a PDF document that isn't attractive, it's named something really strange. And they all come from different insurance companies, different transactions look different. And so, because we email the documents to clients, we have a client portal where clients can see the documents. We don't want them just looking like a string of random letters and numbers. We want them looking presentable of course. So, we've had somebody manually renaming the documents daily for many years. It's hundreds of touches a day, 1000s of touches a month. It's very expensive for us in terms of hours. It's very repetitive, very boring, you have a hard time finding someone that wants to spend their life doing that. So, it was a really difficult pain point for us and it was a really good case I think for automation.
So, when we first met the guys at Quandri, that was the first thing we pitch to them and that's the first process we automated. And as Jay said it's been automated 100% and we've been able to redeploy that staff member elsewhere to make an impact on our customers which is what we ultimately wanted. So that was kind of our first process to be automated. Before we even got through the first launch of that bot, we were all thinking like, what's next? And so, we've now launched bot number two. And the swift digital brand that I mentioned, it's an online broker of course. So, we are spending a lot of our energy trying to get customers to complete online forms and to filter information to us that way about the risks they're looking to insure. So that information is coming into our system by email or by download depending on the application the customer is accessing, that information comes to us and then we were manually transferring that into our broker management system to typing their name, their address, entering their risk information. And so the new bot that we just launched in the last week or so here is actually collecting all that information that our customers are giving us online, it's passing it over into our various software systems and putting in a format that our brokers can then take it and run with it and do the process with the customer.
Prior, I think I haven't done many of them myself but prior, I think our brokers were telling me it's about 15 minutes they were spending taking six or seven emails and typing everything into our system. And then they would contact the customer and then they would carry on their process from there. Now it's really beautiful, actually. They're just getting an activity in their list that says hey, this lead is ready to go. They're opening it up, the files completely built for them, the code systems already built for them and they're just contacting the customer and carrying on.
Another benefit we're seeing huge just in this first week with that is that with online leads and sales in the insurance industry, you have to be extremely fast because after the customer leaves your website, they're probably heading to the next person's website to the same thing. So, we were finding that was taking us 15-20 minutes to notice it was in our inbox and then it's taking us 15-20 minutes to type everything into our system so we were ready to call the customer and it was often like 30 minutes, 45 minutes. If we are A game on like 45 minutes maybe to give this person a phone call. With the bot doing it though, the leads are coming into our system. It's going into our program that we're using all the time so we're seeing the lead instantly. We're not doing any of the background typing for it so we're getting these calls out probably in like 10-15 minutes. And I mean, it's week one but all indications are that our success and closing rate on those sales will be higher because we're being faster with it. So that's another unexpected way and with the automation we had. So those are two bots
Speaker: Jamieson Fregeau 27:24
Great. Yeah and so when you're looking and you're thinking about those thoughts Brock and you're thinking about kind of the benefits that you get from those, would you say the top benefits of those are kind of more financial or is it more on the side of enabling your employees go and work on other things or talk to customers and things like that?
Speaker: Brock Longworth 27:50
Yeah, when I first met you guys actually earlier this year, I was driven financially. I was thinking this is going to be an expense saver and it is. I mean the ROI is definitely there and we look at what we're paying people for the hours they're working versus what we're paying for our subscription to the bot, we are saving money there for sure which is great. But I wasn't actually also considering that our HR director is not spending time recruiting for data entry staff anymore and now she's spending more time recruiting for salespeople. So, there's an added benefit for us. There's a lot of benefits like that that have come up that I wasn't expecting initially.
So now having been months into having a bot working for us though, I would actually say that the ability to take our people and to have them doing things that are more in human natural talents is actually the big win for us. Like we actually have happier people in our data and operations team because they're not just hammering data constantly. They're actually getting to like work on more complex tasks that aren't boring them to tears and that kind of stuff. So, I've actually even noticed like a morale lift for that particular group because I mean, we knew they were in the basement and we knew they were typing data and we knew it was boring but we I don't think they even knew actually how bored they were with it until it was taken away from them.
So, it's been like morale benefits as well. There's been recruiting benefits, there's been definitely the financial benefits that we hope for initially. But I think the biggest win for us to your question Jay actually is that we're seeing that we're providing a better customer experience because we're faster and we're actually able to deploy expert people to customers faster because I don't have a licensed broker renaming an email or a licensed broker typing a phone number anymore. They're actually contacting their customer and they're giving insurance advice which is what we should be doing. So, I think it's just allowed us to redeploy our resources in a more meaningful way for our customers and I'd put that as a number one win for sure.
Speaker: Jamieson Fregeau 29:38
Yeah, that's interesting. And it's interesting to see how that evolves over time of the of the bots running.
Speaker: Brock Longworth 29:45
So, it surprised me actually. I definitely came at it from let's save some money and there was a trickledown effect of all kinds of benefit we weren't expecting.
Speaker: Jamieson Fregeau 29:54
Great, right. Yeah. That really is great to hear and everyone can see why we wanted to invite Brock in. So now that you've seen that and you and you've ran through a couple of thoughts, essentially, what would your process look like for evaluating a new digital worker that you might bring in? Would you still look at those cost savings or would you be looking for a process that essentially improves customer experience?
Speaker: Brock Longworth 30:23
Yeah, so with the first bot the e-doc bot, we actually just work again, coming at it from the money perspective. So, we were looking at it as what is the highest volume click thing that we're doing? So where are we clicking the most basically and that's why we did the e-docs first. As I said, once we launched that e-doc bot, we were thinking what's bot two, what's bot three going to be like? And we chose the commercial insurance lead bots, the guy that's actually taking our leads and putting them into our system. That's not our highest click area but it's probably the area where we have the highest talent or the highest human capital going.
Our commercial insurance brokers have different class of license, they've got more experience and so they're probably our highest valued time and our most insurance experts. So it was just a matter of actually saving their time that was the most important factor for us. It wasn't even really about saving money because we could have hired someone to come in and pump those leads into the system probably but it was just really more about saving time for high value people if anything else.
I think the next bot that we build, probably our main driving factor is going to be where can we free up the most valuable human time and I think that's what will drive us going forward.
Speaker: Jamieson Fregeau 31:37
Right! And what areas of your business or even kind of insurance in general do you think that would be?
Speaker: Brock Longworth 31:46
You know, every insurance broker runs a bit differently. So, I think every broker is going to have their own process, their own system their own way they want to present things to customers. But I think the really cool thing about your process is that it is custom. Like we didn't have to conform to physio(32:02) records standards in terms of how we wanted to have our documents presented. We just told you guys this is what the revisions look like, this is how the endorsement looks and that's how it looks. I think the customization that you guys offer to brokers is really cool. We don't have to look like the broker down the street which is really nice.
But yeah, I mean I think there's any number of areas that we can look at doing that like when we're still having our staff transact payments, we are still having our staff do the accounting transactions on the documents that are downloading. We're still having our staff scan and then manually transact and attach some documents for policies that don't come in our downloads for insurance carriers. So, there's lots of processes that we're still doing quite manually. I think there's definitely another half a dozen bots waiting to go over here for sure.
Speaker: Jackson Fregeau 32:53
I think we've talked about this before Brock like I know when we first started talking and even I think when we first started working together, I know you were skeptical at the beginning of you know whether we could actually deliver that result we were talking about or whether this technology could and there might be people in the audience segment as well. It might be worth it just walking through why you were skeptical, what you were skeptical about and what happened after we started working.
Speaker: Brock Longworth 33:20
I was a tough sell. Yeah, we actually like this the e-doc automation in particular has been a thorn in our side for a long time. And a couple of years ago, I did approach a few companies that were automation or they were working with AI and they were trying to build something like this. So, we actually went down the path and got proposals from some of them. Ultimately, it was cost prohibitive in most cases for us to do it. Like we were talking about enormous development costs and part of it scared me too is that once they developed the software for us, they would give it to us and they kind of just like walked away from it. So, if the download from one of our insurance companies changed, we'd have to recontact them to change it and who knows how long it would take to do that, who knows what the cost would be to change it?
So, there was a lot of kind of barrier reasons why we didn't pursue that. And, we did actually have one of those companies we were trying to work with initially do like a demo for us. And it was really clunky like one transaction from one insurance company in one format. And after weeks of trying it, it wasn't really hitting the accuracy we were looking for. So, we abandoned it, tell him that you guys and your model was just really different. I hadn't seen that before. Like the onboarding cost was really low and then it was an ongoing subscription which was different so it's an ongoing cost for us. But what really attracted me to that was that if something goes wrong, I just send you an email and then you guys fix it. It's not my problem. So that's kind of what I find so attractive about that model. And we have had things that have changed or the format, something comes in changes and we do just let the guys know and within hours, usually it's been fixed and the download is next day and it works fine.
So that rapid response to changes because things do change for us and so that rapid response to change has been really beneficial for us. And that was kind of, ultimately the uncertainty of if things changed and the bot breaks what we do? And that's really what stopped us from going down that path prior to meeting you guys.
Speaker: Jackson Fregeau 35:23
Interesting. Yeah and you hear us use this phrase, “digital worker” and that's really what we want it to be like. If you had an employee running this process and said process change, it'd be easy to go to that employee, tell them to do something different and they would do it, right. And we want these bots to be digital workers like you had a worker there, you can go to that digital worker tell them to do something differently and we can retrain them to work on due process. Businesses are changing, they're evolving, they're always going to be evolving and you want that to be flexible to grow with your business. We do have a question for the audience here Brock for you from an anonymous attendee, what made you decide to go with a partner as opposed to developing it in house?
Speaker: Brock Longworth 36:04
Yeah, so it's very similar. So, as we launched this fifth digital brand, we did build a lot of technology ourselves. So, a lot of our coding systems are proprietary So they're just completely our own. We host them, we built them, we maintain them. So, we have experienced that where we've built the technology, and we've maintained it and, in some cases, it works but something as complicated as a digital worker. I just wasn't comfortable that we had the expertise on staff to maintain it. So ultimately, the subscription model and just offloading the responsibility for updates, upgrades and changes doctor Quandri that was what sold me on this model for ourselves.
Speaker: Jamieson Fregeau 36:46
Yeah, that is great to hear. I mean essentially be going back to your earlier point of changes coming in and the previous solutions you've looked at of just potentially building it and walking away. Because that really does validate the business model we've been working on which essentially, that's why we provided on a subscription and our initial idea of we want these digital workers to be working 100% of the time and working extremely well and businesses are going to change.
So, we want to make sure that we're there and it's built into our system to want to make those changes and make sure those digital workers are working as best as they possibly can.
Speaker: Brock Longworth 37:29
Yeah, I mean, in our industry too, it's not even like we will occasionally change things internally but were also subject to the whims of 12 or 15 different insurance companies that we were with too because if insurance company ABC decides we're downloading things in this format now or we're going to download things over here not over here, that's going to take an adjustment. And if we own the software ourselves, either we have to employ a coder that could change it or we have to go and contract a coder to come in and change it for us and that's not going to be a quick response time. It's not going to be cheap. So, there's a lot of reasons why that doesn't work. And it's not even just change that we can control, it's change that our insurance companies can control and that we have no control over. So, it just it was too much risk for us to not have this subscription model. The subscription model really gave us the comfort to actually go ahead with it.
Speaker: Jamieson Fregeau 38:19
Right, great. So, I guess for attendees here, if somebody was out there Brock and they were thinking about implementing something like this themselves or they might be a little skeptical, what would you say to them?
Speaker: Brock Longworth 38:35
Just do it. Be crazy not to. I think honestly, if you find like a process in your operation that's like really high touch and it's really repetitive and there's not a lot of variation to it whether that's in preparing letters for renewals, renaming e-docs the way that we did it, saving documents in a certain area of your BMS system. There's lots of different things that lots of us are doing that are important but they're just high touch repetitive. Honestly, I think if you found that, talk to these guys, I think you'd be surprised how easily you can be automated. Like our first bots was up and running in like six, seven weeks if even. It was mind blowing the fast to me because I'm used to this technology projects taking six months, nine months and then when you hit nine months, they're saying oh sorry, it'll actually be 12 months. So that's kind of my experience with tech projects.
So, this one launched in six weeks. It was 90 something percent accurate on launch and within I think four or five days it hit up to 99% accurate. So, it was a really fast, really smooth launch. But I think if you talk to them and you find out the onboarding costs, I think you'll find its kind of a no brainer actually.
Speaker: Jackson Fregeau 39:55
Yeah, and just one last question for you Brock before we kind of move over to another section of the webinar here, like what made you want to come onto this webinar? What made you want to share this message with other insurance brokers?
Speaker: Brock Longworth 40:07
Yeah, so typically speaking, I wouldn't get up in front of other insurance brokers and tell them here's how you compete with me. I've been in the insurance industry since I was 15. I've worked with a bunch of different brokerages everywhere I move. I've worked in the independent broker channel and I'm really passionate about the independent broker channel and I really want to see this channel remain viable and competitive in the future. So, I don't really have a problem sharing our success stories if it's going to help another independent broker to compete with the direct writers and that kind of thing. So honestly, I've been going to conventions, I've been looking at webinars like this my whole life and we've always been told you have to get on social media, you have to embrace technology, Amazon's coming, Google's coming, someone's going to come and disrupt the insurance industry and it's a fear that a lot of brokers have that that's going to happen to us one day. But I think that what insurance brokers have the kind of prevents that sort of disruption or is at least holding it at bay is that we're a very much relationship driven business. So, we have to engage with our customers, we have to build relationships with them to retain their business, we have to build relationships to get them to choose to do business with us in the first place.
So I think that ability to build relationships is why we're so people focused, people centric in the insurance industry. But we have become a technology embracing industry as well and it's just a lot of manual process in the back end. So, I think what I'd like to accomplish here is hopefully giving other brokers information that they can use to automate their back-end processes so they can deploy their human capital to do what insurance brokers should do. And if it's a small defense against large disruption to the industry then that's kind of my goal here just to help other brokers to continue holding the disruption for our industry.
Speaker: Jackson Fregeau 41:56
Great, thanks. Thanks for that, Brock.
Speaker: Peter Steube 41:58
Yeah. Thanks, so much Brock. And again, makes it that much more grateful to have you on the line here. I'm curious, I think that this is a pretty rhetorical question. It seems like you do plan on working with the guys at Quandri now going forward and well, you might have been a tough nut to crack initially. Can you maybe talk about where you see that relationship going forward and is this forever friendship or what's the outlook here?
Speaker: Brock Longworth 42:27
Oh yeah, we're friends for life for sure. I mean yeah like the e-doc bot has already made an enormous impact on our business like it really has freed up an entire full-time position that person has been redeployed into our broker support team. So, they're doing memos for brokers, they're doing payment writer phone calls for them. So, it's really actually given us a complete full-time person who is trained in the broker support. So, the e-doc bot is a very welcome member of our team and the commercial leads bots is already saving your commercial team a lot of time. After the first bot, I kind of got eye rolls in the management meeting and I said, we're going to do this but the second bot was embraced instantly and now our management team is talking about what's the third bot and rush getting them coming to us with ideas about like hey, can this process be automated? So yeah definitely, there's going to be bot three and bot four I'm sure and our plan is definitely to work with Quandri for a very long time.
Speaker: Jackson Fregeau 43:29
Yeah, and we hope so too.
Speaker: Peter Steube 43:31
I think there were a few other questions that we had here in terms of as not only for Brock but for you Jackson and Jamieson as the experts who are now pursuing great opportunities across the insurance industry. Where do you guys see your next opportunities to help people like Brock? I know that your services end to end in the sense that you're helpful with both processes defining deployment and then maintenance. Maybe you can talk a little bit about, we covered it a little bit today but what are those next processes that you think the insurance companies are going to be pursuing?
Speaker: Jackson Fregeau 44:19
Jamieson, do you want answer this one first?
Speaker: Jamieson Fregeau 44:22
Yeah really moving forward. Currently, we have the digital workers that we really believe make an enormous impact in customers business like Brock. And essentially, for us, we're looking at how we can take these digital workers and just enable more people to access them really. So, we're working on a few things internally that we think are pretty cool and pretty exciting that should be deployed within the next couple of years, within the next year really all for the goal of making these digital workers faster to deploy and more accessible to brokers.
So, an example of that is with the e-doc bot itself. When we first built this e-doc bot for Brock and Cornerstone, we needed to build in and train this bot to understand the 10 different insurance companies, their policies, what they look like and enable a bot to be able to digest that and understand the numbers on it. But now that we've done that, our digital workforce essentially understands all those policies. So, moving forward, we don't need to spend the time to teach the digital worker to learn those. And essentially, as we continue doing this and building this technology internally, it will just enable us to really perceive digital workers out faster, more efficiently and really be more effective and pass those cost savings on to the customers really.
Speaker: Jackson Fregeau 45:56
Yeah, good answer, Jamie. To add on to that, I would say we we're building these bots right now and we're leveraging a few different technologies. We're using RPA, we're using OCR, we've used a little bit of machine learning. And as this progress and as we come across more bots, we will be doing more use cases, we'll be pulling in more technologies, we're going to be leveraging more artificial intelligence. So, these bots are a little bit more independent and can make some bigger decisions. And ultimately, what that means for brokers is that we're going to see a really big expansion of the use cases that these bots can do. It'll go from doing complex full automations but basic decisions throughout to bots that can really handle much more complex decision logic and can handle a lot more use cases. And as we do more use cases, what we want to be able to do is be able to launch those automations quicker and be able to launch them cheaper. So, there'll be a wider range of use cases that we can deploy much quicker which means that you can realize these efficiencies a lot quicker in your businesses.
Speaker: Peter Steube 47:00
Excellent, excellent. And we are approaching the final 10 minutes here and there's also been a little bit of a lively exchange here in the Q&A portion. I'm not sure if everybody's seen that but I'd like to call out maybe a few of the more poignant questions that have come in here over the course of the webinar. One of which is are these robots running in the cloud or are they running on Brock's desktop or on his premise. Jackson and Jamieson maybe you can talk a little bit about that deployment capabilities and this is where Robocorp will come in as well in terms of enabling with through that technology and that distribution of the bots as well.
Speaker: Jamieson Fregeau 47:37
Yeah. So it essentially depends on the bot that you're building and what that process is. So, for example, the e-doc bot is running 100% in the cloud. And essentially because it's just accessing epic(47:53) through the browser and downloading those attachments and then renaming them and uploading them. So that's running completely on the cloud independent of any on premise server or anything like that. But for other bots, sometimes you need to be able to access a desktop application like policy works for example. And so, for those, then we spin up servers and run those processes on those servers themselves. And yeah, as Peter said, that's where Robocorp comes in essentially enabling us to easily and not have to worry about the underlying server technology, deploy those bots onto the cloud or run them on a server itself.
Speaker: Peter Steube 48:39
Excellent! And then I think a few of the other questions that I want to highlight here is how long did the entire process take right. So you have a process defining and then building and then testing the process, make sure you know, how quickly was Quandri able to deliver here?
Speaker: Jackson Fregeau 49:02
Still not used to these video calls. From end to end like from a contract signing to that digital worker actually being deployed and working in a live environment, depending on the complexity of the process, it can take anywhere from four to eight weeks, somewhere in that range. I think for this e-doc example that we've been talking about, I think it was deployed at about the six week mark and we go through three main phases whenever we're building bots. There's the design phase which can take from one to two weeks and that's where we work with a broker's team to define all of the decisions and the logic of that process. And that is really one of the most important parts because we then take that and that's what we pass on to our development team to actually build that bot. So, it's really important that we get all of the nuances, all of the decisions, all of the exceptions, all of the business logic in that design phase.
The development phase, that can take anywhere from two to three weeks and then finally, there's the deployment phase where we run a week of doing mass testing to make sure that when we launch that bot, that bot is working from day one. So those are the three phases that we go through to launch any bot.
Speaker: Peter Steube 50:11
Thanks for that. And actually, I'm very curious Brock, from your perspective, what was the resource commitment on your side right? Like do you needed I assume yourself or another business leader to at least document these processes. But how helpful was Quandri along the way here? Did you feel that was a little bit of a 50/50 effort or was that you really felt like a white glove treatment?
Speaker: Brock Longworth 50:35
Definitely a more white glove. Yeah, it was like we would have an initial meeting, we would kind of outline the use case that we were hoping to use the bot for and then we'd set up another meeting where we would actually just kind of screen share and we just would walk them through the process for one of our people was actually doing what they would do with that process. They would record it and then go away and come back to us with that handful of questions that one of our people who is familiar with the process would answer and then we would feed them some samples of the work the bot was going to do, they'd keep working on it. And then we'd ultimately launch the bot and the first week or two either they're manually checking everything the bot does to make sure that it's accurate. But yeah, from our perspective, it's really just almost letting them like watch our process as we're doing it a little bit and then just answering a few questions over the next while they work on it. So, it really wasn't a huge resource commitment on our side.
Speaker: Peter Steube 51:31
Right. Great! That's great to hear. And then we have another question kind of back to the Quandri guys. So, we talked earlier in the call about the differences between UiPath and Robocorp. We're picking on UiPath because they're the 800-pound gorilla in the room but there is an entirety of a gen one providers or solutions. There's basically a trilogy of them.
But anyway, can you talk about have you done actual replacements of existing bots from other vendors? And one, is that doable and two, is it easier, is it harder? Kind of comment on your experience doing that.
Speaker: Jackson Fregeau 52:06
Yeah, definitely! We've actually done one of these. It was earlier this summer so it's a fairly large one in the month of July. They weren't actually using UiPath but they were using the other major large RPA provider which is Automation Anywhere. So, this company had a bot running on Automation Anywhere and the main problem with it was that it just wasn't really working. They've been running it on there for I think a few months and they were consistently errors coming up with it and it never really worked as expected. So, we got brought in to actually transition that bot from automation anywhere onto Robocorp. So, it is very possible and I would say it's slightly easier than starting from scratch because you have all of that decision logic already built out. So as long as that business logic is correct, it makes it easier because then you can skip that phase of the process but you do need to move very slowly through it because it is building blocks of code which is kind of difficult to understand so you need to be very careful of what you're pulling over. But we can take that code from either your pseudocode from either UiPath or Automation Anywhere and we can translate that into a live bot that's running in Robocorp and we do that in Python.
Speaker: Peter Steube 53:26
Excellent! Thanks for that. And let's see if we can maybe pull some other questions that came from the audience. There is a remaining one about AI and potential integrations or uses there can use. Talk about if you have any preference in frameworks or is there been stuff that has really caught your attention?
Speaker: Jamieson Fregeau 53:55
Yeah. So essentially, I guess leaving OCR aside because we do use quite a bit of OCR to essentially get text and convert that into something that we can use from images or occasionally a converted PDF into an image. But yeah, so we play with TensorFlow generally for a lot of our artificial intelligence implementation and essentially, where that comes in is that large scale data, understanding large scale data. So that can be beneficial for is essentially, if you have this bot, analyzing a large portion of data and there are missing or potentially incorrect values is one use case we've looked into. Essentially you have all these data values and they're being entered by people and occasionally something is not there. So, something we've implemented is using artificial intelligence essentially to predict what some of these values are most likely based on the other values in that table of data. There really are a lot of different use cases for using artificial intelligence. And we really see RPA as essentially being able to be the hands and the feet of artificial intelligence of you have these algorithms making these decisions but instead of a person going in implementing those, you can use RPA to go and implement those changes but also gather the data for the algorithm to learn on. So, there are a lot of really exciting things you can do using that space.
Speaker: Peter Steube 55:43
Excellent! And that's is bringing us right to the final two minutes of the webinar and we're always conscious of people's time that they've contributed to us already. But is there anything else that we missed today. I'm going to leave the floor to the smart innovators on the line Jackson, Jamieson, Brock, is there anything else that we missed besides obviously to get in touch with you guys and your phone lines are open. But thanks, so much guys for joining this webinar. Thanks for being great stewards of Robocop technology and Brock, thank you for your contributions to the insurance community as well. I know that we had a quite a few number of your peers on the line here and I will say thanks on their behalf. So, leaving it to you three guys to wrap up, if we missed anything, feel free to go ahead.
Speaker: Jackson Fregeau 56:31
Peter, thanks for organizing this and thanks for having us on. Of course, it's great to come on and talk about what we've been able to do and of course to bring Brock on, Brock, thanks. Thanks for doing this with us as well. The only other thing I'd like to touch on there, looks like there was one more question from Andre. Can end customers go and use Robocorp directly? Absolutely! You don't need to go through a partner if you don't want to. You can go use Robocorp. If you want to build that bot on your own. If you want to manage that bot on your own, you can absolutely use Robocorp. Where we fit in is building those bots and then managing those bots are ongoing basis. We have built some bots that we don't manage. So, we are flexible with how we work within the Robocorp model as well. And really, it's up to you and what works best with your business.
Finally like we'd love to hear from anybody that's interested about how this technology can help your business. So, Jamieson and I we've blocked out time over the next week to make sure that we can talk to anybody that is interested. So, please reach out to us or go to our website if anybody wants to talk about how we can save you time.
Speaker: Jamieson Fregeau 57:35
Yeah, great. And yeah, once again, thanks Peter a ton for having us here and thanks Brock for essentially giving the inside scoop on the insurance industry. For us. That's super helpful and thanks, everybody for coming. And even if you don't necessarily have a specific process that you think would be applicable, it'd be great to hear from everyone about essentially what you think that this technology can do in the insurance industry and here are your two cents on it. So, reach out to us and we're looking forward to chatting with you
Speaker: Peter Steube 58:14
Thanks so much. Anything to add in final thoughts Brock or are we good to get back to that time that you've been reenabled through the bots that Jackson and Jamieson have created.
Speaker: Brock Longworth 58:27
An awesome time the bots are doing at all. Yeah, I think it's good. Honestly, I would just really encourage other insurance brokers to be willing to explore this technology. It's something that I was a bit timid to do initially as we alluded to but it's definitely been a huge benefit to us. So, give it a try and inaudible (58:45)
Speaker: Peter Steube 58:44
Great! Thanks so much, guys. Well, enjoy the rest of your day. I know that it's earlier for you guys on the west coast of Canada but for everybody else on the line, thank you again for tuning in. There will be a replay posted very shortly after this. And of course, get in touch with Jackson and Jamieson right away if you are interested in learning more. Thanks so much, guys. Enjoy your weekend.
Speaker: Jackson Fregeau 59:08
Thanks, Peter. Thanks a lot, everybody.