Ron Glozman
Ron Glozman from Chisel AI discusses how Chisel AI is focused on automation for policy checks, quote comparisons and more. Ron wants to enable those in insurance to get more work done sooner so they can have more time to spend with family and friends.
The full transcript
Transcribed automatically from the episode audio, so some names and terms may be misheard.
James BenhamJBKnowledge Podcast Network. On episode 42 of the InsurTech Geek podcast talking about AI solutions for commercial insurance with Ron Glozman from Chisel AI. The InsurTech Geek podcast powered by Jimmy Knowledge is all about technology that's transforming and disrupting the insurance world. We'll be interviewing guests and doing deep dives and specific techniques you see changing the industry. We're taking you on a journey through insurance tech, so enjoy the ride and geek out. Back in a second. Alright. Alright. Alright. Another day. Another InsurTech company to talk with, talk about all things technology and insurance and all the great stuff that's transforming this industry that we all love.
And today with me, I have the illustrious, the Canadian a, Ron Glozman from Chisel AI. Ron, good to have you on the day joining us from Canada. Pleasure to be here. Thank you for having me, Yeah, Sorry. Let's have some poutine and and call it a day, I I never understood what people saw in poutine. It's it's not for me and I think a lot of Canadians will will say I'm not Canadian for that but Yeah. They'll they'll disavow you, It's it's a it's an interesting deal, Ron.
I'm a little out of sorts right now because I usually go to Canada at least two or three times a year. I like going to Vancouver. I have some long time conferences I've been speaking at in Vancouver. I like going to Toronto and speaking at a couple conferences there every year. I generally do the rounds. I hit Calgary, Toronto, Vancouver. I'll go to Quebec City. None of that this year. You know? I'm banned. I'm personally prohibited. They don't want me there. Oh, wait. No. No. It's because I'm an American.
They don't want any Americans there. You're back in Code Red. Code Red. I feel like we're on Star Trek now. Captain Putter in Code Red. You know? Like, Justin's like, put it in code red. I mean, jeez, Louise. You're you're on you're you're you're shield your forward shield's maximum. Right? Like, you're like, we gotta we gotta stop all phaser fire from south of the border. Keep all the virus out. So I can't come see you. I can't come hang out in Canada. How are things up there as winter as winter arrived?
Ron GlozmanIt's funny you say that we had our not our first. It's it's weird because we sort of have 13 seasons in Canada. I know you've probably heard of four seasons, but we have fourteen fourteen here. Here. And one of the seasons is called early, early winter, and then we have late fall and then we have real winter. So it just snowed yesterday. We have a thick, nice covering on the ground, but that was the second snowfall of the year, but I think this one will actually stay.
Well,
James Benhamcongrats, I guess. What let the winter sports begin? I mean, in code red.
Ron GlozmanLike, you get to go out on your snowshoes and now your snowmobile? It's funny. I asked somebody the same thing. I'm a I'm a very avid snowboarder, and I was thinking, you know, chairlifts are six feet apart in some sense, at least when they're going up. So it shouldn't really be a problem. So I hope that they open mountains, but I don't think there's been any official statements.
James BenhamYeah. Well, I hope they do because, man, that would be a shame. I'm a I'm a skiing guy, you know. Snowboarding involves too much time on my butt. I like I like I like skiing a whole bunch, but I hope you do get to do your skiing. I know you Canadians, you thrive on winter sports and visits to Florida in the winter. So since you can't do visits to Florida, hope you at least get your winter sports. We're not going to talk about the weather though.
We're going to talk about insurance tech. Before I get started with my interview, don't forget you can subscribe to the InsurTech Geek podcast. If you're listening to this on, YouTube or Facebook or any of our live video streams, you can text geek out to 66866, and you never miss an episode. We'll email you the show notes and the links and everything there. Back to Ron Glozman from Chisel AI. That's Chisel dot a I. Ron, we're gonna talk about Chisel and what it does. I'm pretty excited about using machine learning for, you know, making sure people are actually getting what they bought, you know, and they got the quote.
I mean, that's that's really as a as a consumer of a lot of insurance, I can tell you that's a big pain in my rear end. So we're I'm excited about that. We're gonna talk about how you're helping underwriting too. Before we do, I wanna talk about you've got a good story. You're Canadian, which which is cool. Right? I mean, you got maple syrup in your blood, don't you? And you so tell me, where'd you grow up? What'd you dream of doing when you were a kid?
Yeah. It's
Ron GlozmanSo I actually immigrated to Canada, which I think is the most Canadian thing. It's I was I think that statistically, like, two thirds of people who live in Canada are immigrants. Although I might be wrong. It might be one third, but still, it's quite a large It's lot. Yeah. But I came very young and I obviously loved it. I From where? Always told as an immigrant. I came from Israel from just outside Tel Aviv. And my parents always told me you gotta be a lawyer or a doctor.
And I think that's a very common joke, but it's it it it actually did happen. And but they also said that one more option or an entrepreneur. And I think they actually highlighted to me that because I I would often as a young kid, would say, at Bill Gates, look at Steve Jobs. At the time, it was more about money because I thought that was the most important thing. And and since then, I've learned there's a lot more things that are important than just that. But I said to myself, you know, I wanna figure out a problem that I can solve because at the end of the day, that's how you really be really become successful in more ways than one.
But I had no problems to solve. And when I was 14, I I came across sort of my first idea where I said there's a lot of piracy out there in the world. Statistically speaking, 19 out of every song downloads in the Internet in two thousands were illegal. So if they were making billions of dollars on 5% of the market, how much money was out there for the other, you know, 19 x that they were leaving behind? And so I said to myself, what if we do an opportunity where people can listen to music for free and we make ad revenue?
And it was pretty successful. We had a couple thousand artists, like about a 100,000 users and broke even on the venture. But obviously we now know Spotify has obviously come in and really dominated that space, but it was my first learning opportunity and I learned a lot. I sat around, went to college at university of Waterloo in Laurier where I studied computer science and business. And I said, I said to myself, you know, I was sitting there as exam season. I said to myself, it just doesn't make sense that people have to study by reading textbooks because exams are only 10 or 20 pages and the whole textbook is a thousand pages.
That's one or 2% of the content. And I said, I think I can teach a computer how to read because that seems pretty fun and easy. And so I spent a semester working on it and I got it to the point where I could stop going to class. I could just, you know, take the textbook, run it through the summary app, study for four hours, take the exam, and still get an eight. Good.
James BenhamCome on. Then your university found out about it and they made it illegal. And
Ron GlozmanWell, actually, you would think that, but I actually went to them. I tried to pitch them and they gave me $25,000. That was that was a good start. And I also burned out on the app store and it ended up going viral in about two weeks. It was in 33 countries, 44 of the top Ivy league schools in the world. We're talking Princeton, Stanford, Harvard, Yale, Russia, China, India, Brazil, Portugal, Switzerland, Netherlands. Like, you name it. It was everywhere. And it went on to be named one of the best 50 apps for students of all time.
So Nice. Really, it had some success. And you wrote and you wrote it a 100% yourself? I did at that at that point a 100%.
James BenhamYeah. So what'd you what'd what'd you use to write it? I'm you know, I'm a nerd, so geek out with me for a second.
Ron GlozmanSo here's the interesting thing. This was, I would say, the first year that textbooks were really in available in PDF format. Not oftentimes, you know, they used to just be print. So having to scan it and have to having to do OCR was a big challenge. So one thing that I got to really leverage on was the ability to read native PDFs, so I didn't have to do OCR. What I started with my as my training corpus was oftentimes when I would read textbooks, I would highlight in yellow the things that I thought were important.
And so I would go through and I would use that as actually my training set to say, okay. Here are the things that I think are important. I then started to incorporate additional factors. All of it was built in Python. Maybe let me start with that. All of it was built in Python. I used a lot of libraries like obviously pandas and NumPy. One of the ones that I'm a big fan of that at the time was not very famous, maybe it's well more well known now is called spaCy.
Yeah. Really, really big fan of spaCy. And basically, I ended up building a lot of tools around TFIDF, around, you know, tokenizing and stemming and, you know, word to VAC and glove and all of these things. And at the end of the day, I was able to take a textbook and summarize it to one page for every chapter. And then if I wanted to, I could of course customize that to be X amount of pages long. That's
James Benhamawesome. So you did it. You did, you know, you were studying two degrees in two different colleges at the same time. And at the same time, you're you're writing, some machine learning algorithms to bypass all the all the the liturgy of thousands of pages of reading. You're acing your tests. What happened? What, you know, something happened, right? You have the entrepreneurial spark, you realize, Hey, I should be doing this for money. What's the origin story?
Ron GlozmanYeah, it was very roundabout. As I said, it started as an app for students and I worked on it in that form for about two years. And in early twenty sixteen, late twenty fifteen, was invited to a machine learning conference where I was asked to present on a panel about natural language processing. And about five minutes after I got off stage, an email came in from our central inbox and on our homepage. And it just said, Hey, just saw you present. I know this is an app for students, but I think insurance can really benefit.
Do you have five minutes to chat? And it ended up being one of the biggest insurance brokerages in the world. I quickly followed up with her. She was a senior VP there. Said, we'd love to chat. Met up with her and her team basically six times over the next eight weeks. And at the end of it, they educated me a lot about the types of documents that they read when they're trying to avoid errors and omissions in their operations. And then they said, do you think you can teach a computer to read and do the same job?
And I said, yes, I think I can. I think it'll take, you know, six months. We negotiated a price at the time. I thought it was a crazy killer deal. I knew that there was a market there. Even if I hadn't done the traditional Tams sizing just yet. And off, we went to the races. Off, we went to the races, meaning you actually went live with that broker. So we did a POC with them. This is a large multinational broker and we were starting in Canada.
So the next phase was then to go to The U S and convince them. As you know, sometimes that takes a lot longer and requires a lot more manpower and a lot more trips than you can expect. In the meantime, we actually went across the street to their biggest competitor and sold it to them and went live with them.
James BenhamSo the one that gave you the idea wasn't even the first client. Correct.
Ron GlozmanNice. We were the first client. Weren't the first one in production though. Well, I mean, you could say it's the same thing. That's awesome. So
James Benhamwhat was the outcome? Because I I like talking about, like, ROI and outcome when it comes to tech. Right? I've been writing software like you since I was a a early I started when I was 11, but I've been writing software a long time, love talking about tech, love building tech, but at the end of the day, it can't just be tech for tech's sake, it's got to actually make money for people. So, the broker that actually went live with it started paying those guys. You don't have to say who they are, just how much money did it how much money did they spend versus how much did they say?
Don't have to give their their spend, but, like, you know, how quickly did they make their money back? Was it, like, two or three months? Was it a two or three weeks? Two or three days? Like, what what are they telling you on on ROI and payback? Yeah. It's interesting. It really depends how you cut it. So I'll I'll give you a number. I'll say thirteen months.
Ron GlozmanI also think it really depends how you cut. And I think thirty months is good. I think that's a pretty good ROI on a project. And also the question is, you include fixed costs? And I would in my sense, fixed costs would be the professional services. You know, when you do B2B enterprise, a little different than B2C, you typically have a services component. So sometimes the services component in year one can be almost as large as the SaaS fee. And so if you don't include that, the ROI is much faster, but if you are including that, it's typically like thirteen months.
And so what that means is that can literally be like 10 times cheaper than it actually costs them today. 10 times cheaper. Than it costs today. So just as an example Then it costs for them to do it manually today because they're doing this, they're doing this
James Benhamreview and in this particular case, you're making sure quoted is what was bound. Right? You're Exactly. You're reading the quote and then you're reading the binder and then you're comparing them and you're looking for additions, additions and exclusions, additions and removals, which happens all the time. It drives people crazy. It also makes people look dishonest. Right? Right. Like, hey, this is not what you quoted me. So they're doing manual review. They go to do automatic review. I'm I'm guessing they do just some manual double checking on top of that, but the the total labor savings are huge.
Right? That's that's right. Like we're talking 80%,
Ron Glozmanlike in the range of 80%. Incredible. Incredible. What year was that first go live? Oh my God. Well, it took a long I would say officially first go live was at the 2018, 2019. Yeah. So we're coming up on two years now. Yeah. You're coming up on two years. So how has the account expansion gone? Yeah. It's interesting. We since then, we've had a new client come in who's been even bigger and, you know, think I this one's a little more public so I can talk about it.
We, we won the Zurich innovation world championship last year, the inaugural year. Nice. And fun fact, I'm not a big golf fan, but for those of you that are Zurich hosts or sponsors, one of the big golf tournaments and they keep the trophy for it. And Tiger Woods is one of the trophy holders and my trophy from the Zurich world cup sits right right next to his in their main office in Switzerland. That is awesome. I think that's pretty cool. Anyways, back to the story. So Zurich's been very public about their relationship with us.
That is one of our, you know, most sort of favorable relationships, one of our largest accounts. So I I would say you learn a lot. Obviously, the first customers let's talk a little bit about mistakes. Like, the first customers are the ones who you learn a lot from, and sometimes you make mistakes with them that you live and you get better from, and sometimes it just takes them time to see you in a different light. And so sometimes the expansion can be faster with other customers.
James BenhamYeah. Okay. That's great. So where have you evolved to? So you're not just doing machine learning review of binders to quotes. Now you've dive, you've taken a dive into the deep end of the pool with underwriting, correct? That's right. The deep end. That sounds so scary. But it is the deep end. Underwriting is a big, deep, scary place. Yeah.
Ron GlozmanAnd I think that's interesting because it's, I would say it's not a deep end of a pool, but it's like Mariana's trench. Like there's so many levels to this. And oftentimes as a startup, even as investors or even with customers, it's hard to differentiate and they'll be like well there's a million people who do this and it's like well not really because like there's levels there's deep there's levels to this but yes so what we're starting to do now is focus on the clearance and registration piece of underwriting.
So the interesting thing is when we look at underwriting there's depending how minute and granular you want to be fifteen, twenty three or 42 touch points. You'll probably get even more than that but at least in our analyses we've been able to get those numbers. So when we think through that, one of the very first things that need to happen and it's one of the most pivotal things and the things that companies have the worst turnaround time on is clearance and registration. And what I mean when I say clearance registration is typically three things need to happen.
An application or a submission comes in from a broker or some other method. Sometimes it can be through a portal if it's still a PDF. And what typically happens is there's a queue. So it sits there for three to five days until some type of customer service representative or underwriting assistant can get to it. Their job is to look at the submission and just do simple data entry, like get into our CRM, get it into our clearance and registration system to check for sanctions and get it into, you know, let's say our quoting engine for some other reason.
The other thing that they do is they also try to decline the business. So if they see things like, Oh, these guys have like 20 claims in the last five years, that might not be a risk appetite. So that determination can take time and it also has a lot of mistakes. And so the very, very first thing that we focus on is just getting the information out from those applications into your traditional systems. So it's meant to work with your Guidewire system, your Salesforce, whatever it might be.
It's automated entry. So you'd use an RPA for that? It's a lot more complex than RPA because here's the thing. A lot of these carriers will work with more than 3,000 brokerages and agencies, all of which have a different format. Not everybody's using a board. Yeah. Don't you love it? It's like, it's like an IT, it's like a technology
James Benhamdream because it's the, it's a gigantic colossal problem. Right? It's an incredible amount of inefficiency and absolutely no standards. It, like,
Ron Glozmanit screams out machine learning. Right? That's right. And that's the thing like with RPA and it's also interesting because Well, there's some RPA bots that have machine learning built in that do a really good job of adapting though. That's right. And I think also the interesting thing is if you look at the distribution of the problem, most of your business probably only comes from a handful of those brokerages and agencies. And so even if you tackle head, it typically has a short head and a very long tail.
Yeah. But that short head is very fat. You might be able to solve sixty, seventy, 80% of the problem by solving ten, twenty or 30 different file types. So RPA can definitely leverage the problem on the, on the small scale. But when you get to really, really big problems, it doesn't quite solve the problem the way we do. How do you do it? Oh, baby. Like we have a pipeline. It's not like one piece of code. It's like, it's very like
James BenhamYou just copy and paste code off of a form stack or something?
Ron GlozmanI mean, you definitely do. Stack Overflow. You definitely do, but that's a small piece of it. No. I mean, I won't go into necessarily
James Benhamthe super, super saucy stuff, but I'll go maybe You don't have to. You can keep it at 50,000 feet. I mean, this is because I feel like we don't get enough time to talk about this. Like, we I I get it. I understand what you're So first off, you're trying to you're trying to make sure that forms are complete, right? That has to be step number one. Like, an application in that in that the the first check, is this application even complete? It's interestingly enough. Yes.
The answer is yes, sort of, but it doesn't take any action. It just says the application's incomplete. Yeah. It just lets them know. It didn't it didn't kick it back. Kick it back. Yeah. It does not kick it back to the applicant. It just lets the
Ron Glozmaninitial person dealing with it know that it's incomplete. Correct. And obviously, like that's a configuration thing. It's very easy to write software that says once this flag is triggered, you know, send an email, but no, so basically the goal is that the emails come into a centralized inbox or something else. The inbox is constantly pulled sort of like a web hook. It's constantly monitored as applications come in. The first thing you do is you read the email body. That's one piece of data. You extract some information from there.
Typically, let's say the name, the email of the broker, the name of the applicant and the date that they need the response by. Then you need to look at all of the attachments. And on average, there's three to seven attachments. I've seen as many as 33. So then you need to take each one of those attachments. It can be word files, Excel, PDF, and each file has to have its own parser or its own model or pipeline that it goes through to be read. So you run each one of those, you extract different data points and then you need to know like, you know, let's say there's conflicting information Like this says, this is the phone number.
This says, this is the phone number, which one takes precedence. Yeah. And so like, it's a deep, deep, like, it's I call it a cake. There's levels, many, many layers to it. That's the 50,000 foot view of how you do it. That's awesome. So what what's the market reaction now?
James BenhamLike, is it, oh my gosh, I have to have this now. Are there still a lot of people that don't understand what you're doing? Education is definitely a P I would say now it's different. The education that we had,
Ron Glozmanyou know, four years ago was a lot about like what is AI? How can I trust it? And a lot of preliminary education pieces around what is AI? What is technology? Now the conversations are a lot about how are you different than your competitors? Educate me on how you're different. So they're no longer saying is AI the right solution? You know, what is AI? Is it safe? But they're saying, how are you different? So now we have to, and back then I would say we were the only vendor and nowadays I still think we're the only vendor in specifically what we do.
Everybody thinks that by the way, we don't have an, we don't exactly have a head to head competitor. Yes, you do. Right? Like everybody does. Yeah. Yeah. Course. And there's a couple that come to mind, but I would say a lot of the education nowadays is, is more about competitor versus like, what is AI and how does it work? Yeah. Gotcha. That's awesome. So what's next? We want to build more applications and I think we have a couple of really interesting areas that we want to work on next.
And these applications are, I'd say today we're very much in laying the plumbing or the groundwork for being able to do cool things downstream. So right now, as we talked about, it's a lot about making sure the data exists, making sure the data is clean, making sure the data is valid, making sure the data is in the right place. Once you've cracked that nut. And I would say a lot of companies need that nut cracked and nobody has, you know, or people are still working on cracking the nut.
Once you get that done, you can move on to really cool things like what Amazon's amazingly good at, which is like, what do other consumers like you buy? Like imagine an insurance company that could have the recommendation engine. And so that capability is something that I would love to have a part in building. I think the first person to be able to do that will reap the benefits. I think Amazon is also trying to do that. I read in The US, if you're not buying drugs, like pharmaceutical drugs
James Benhamwith insurance, you can get it at a deep discount on Amazon Prime. Yeah. Well, there's there you know, we have a very complicated medical market here, but Amazon is jumping in headfirst into all kinds of segments of it. And, of course, Walgreens and and CVS are as well, They're pretty direct medical care. So it's a it's a pretty it's a pretty different world that's coming up. That's
Ron Glozmangreat, man. Well, look, let's just kind of the last thing. Where do you think machine learning learning is going? Is this is this replacing humans or augmenting them and making them better? Or a little bit of both? Both and, like, even more, like, it's also creating new jobs, like data science jobs are some of the most in demand and highest paying jobs that people can really get and There you go. Coming out of college. Right? So it's like, it's a complex question. I liken it to like the industrial revolution where, you know, a lot of people used to be employed in the agriculture industry and nowadays 4% of people feed 96% of the population.
And so I think we have sort of a similar thing where we'll be able to reap the benefits. We'll be able to have better lives. Hopefully in some sense, I'm hoping work less. So one of our company missions or our company mission, our company saying is work smart and enrich lives. It used to be work smart, not hard, but we decided to change it because, you know, not not a great saying in the sense, you know, you should still work smart and hard. Yeah. Know. I I
James Benhamlearned that a long time ago. I started my company when I was 21. Remember I when I was 25 and running a tech company. Right? It's it's been been been a while. I'm 41 now. But I used I I used to always correct people when they would say work smart, not hard. I would say, how about work smart and hard? Because Right. You know, like we we should work hard every day. It doesn't mean to work long hours. You just have to work hard during the hours we're working.
Right. And that's right. And that's that's to me, that's actually the biggest promise of AI and machine learning is allowing us to keep working a good hard eight hour a day, but to get a lot more accomplished, drive output up, drive productivity up, which drives, you know, earnings and you said exactly the jobs that are getting created by machine learning pay way, way more because they add a lot more value. And if you look at just over in we're functionally in what the third industrial revolution, right?
I mean, the first one started two hundred years ago. Each time unemployment net aggregate unemployment's gone down and wages have gone up. And, and that's net of inflation, right? So like just raw wages and then, and we're, we work 50% fewer hours than our great grandparents did on average and that came from automation. And, and so when you look at that, I mean, the five day work week is less than a 100 years old, automation. There's a lot of things that came from automation. There were certainly legislative reform that came along with it, but that legislative reform wouldn't have been possible without productivity gains made by the industrial revolution.
And to me, this is the third wave, right? This is digital industrial revolution. Although there's still, you know, rapid changes and improvements being made in raw manufacturing. I was looking at some videos of some some modern car manufacturing facilities, and they are absolutely science fiction brought to real life when you see them going on, you know? And then you watch, you watch SpaceX launch their crew dragon capsule, a Canadian who went to South Africa and then back to Canada and then came down to the promised land here in The US.
Elon Musk? Yeah. That's right. Yeah. A lot of people don't know that. Yeah. Yeah. Yeah. Why know that. Yeah. His his Canadian family, a very adventurous Canadian family who decided to go live in South Africa for a while. Then he came back to Canada first, didn't like the job opportunities he had. So he came down to The United States, then, you know, applied to Stanford, never enrolled, never actually enrolled at Stanford. And it's it's just it's amazing to me how much value has been added to humanity through engineering.
Like, it it I I honestly get a little emotional about it. I'll be straight. Like you and I, you and I share a lot of lot in common. You know, I, I did anything I could in college to try and game the system. Right? I was honest. I didn't cheat, but I remember doing side projects for clients because I started my business when I was in school and I would go to my professors and get them to let it count as my work for the class.
And, and like, I, I got in really big trouble. My junior year, I was an, I was an ROTC. I was a, I was a Navy midshipman. So I was in the military college portion of Texas A and M. And I got in a lot of trouble my junior year for this event that we weren't supposed to do, and we did anyway. And and, you know, I got a rebellious streak in me, and and I was the senior ranking I was the senior ranking junior on the scene.
And so they came down hard on me. And so I went to the head of discipline discipline, and I said, I noticed that you're tracking all of the disciplinary items on, paper. I said, would you like to automate that so you can automatically know how much, you know, how many penalties have to be assessed on people? And he goes, yeah, that'd be great. Said, okay. How about I build that system for you and you let me out of my restricted weekends where I can't go anywhere?
And he's like, deal. I was like, this is awesome. I mean, it just it's just so it's just so funny. And but but the really cool thing is every one of those projects that you work on when you you know, and I'm sure you've worked on a bio I looked at your resume. I mean, you you had you had jobs all through college. All those projects add so much value in whatever whatever field they're applied. And you look at this, the big problem that you solve to me when I when I I got really excited when you agreed to do our our show because it's that you're helping people actually be properly insured.
This is literally one of the biggest problems facing people who buy insurance is that what they think they're buying is rarely what they actually get bound on. And you've functionally fixed it. And that's my So it's like you have this big moral good and then you have this big business good, right? And the brokers save a lot of money, the carriers save reputation face at the end of the day, they really do, whether they believe it or not.
Ron GlozmanMost importantly, the person buying insurance has the coverage they needed. Yep. And to the question that sort of kick started this, do I think it's gonna impact jobs? I think the interesting thing here's what I've seen. There was a gentleman, you know, one of our early customers that I went in and I did a time study and I shadowed him for a day to see what a day in the life of a broker is like, you know, riveting stuff. It actually was actually very exciting. And I said to him, you know, tell me about your day.
And he goes, usually come in, start around seven or 8AM, and I'll leave the office around seven or 8PM. And I go, holy cow. Like, why do you work twelve hour days? And he goes, I need to check four policies a day. And when the policies take two hours each, that's a good day. And when policies take four hours each, that's a bad day, but I need to check four every day. And so by helping reduce the amount of policy, the amount of time it takes him to check policies, you know, my vision is he's gonna get time back in his day so that he doesn't have to work extra hours.
He gets to go home to his wife or to his partner, have a great dinner, go to that baseball game with his kids, whatever it might be. And so I think that's also another aspect of sort of the greater good. It's not always about taking away jobs. You know? Yeah. Yeah. I it's never
James Benhamin in all the discussions I've had with every technology company I've been involved with and every piece of code I've ever built, we never say how can we eliminate jobs. I've been doing this twenty years, man. I've been writing code for thirty. I've never uttered those words. That's right. I love that. And because that's not what it's about. I like people, man. I'm not an introvert. I'm not like a hermit. I like hanging out with people. I like humans. I want people to be able to do things that matter.
And I think you do too. Like, all right. Would you rather have a job or you didn't even in college, you didn't even want to sit down and read the textbook because you knew how pointless it was, right? And so imagine these poor people who have to just sit there and just read these volumes of documents, just making sure someone didn't screw up a binder. That's right. And it's like that there's more to life than that. You don't want people sitting in a job going, there's more to life than this job.
Like you want to say this is fulfilling professional work, right? Yep. So that's what I say when I go to write code is how can we build something, not that eliminates jobs, but that eliminates meaningless work and allows those people to do meaningful things for those companies, you know? And I think that unfortunately there's people out there who really hate technology and you know this, you're 25, you've been on this earth long enough, you've been in quarter of a century, you, you know, you've, you've run into them, right?
They just, they, they say it's evil, it's eliminating jobs and we got to tax it. I mean, stuff like taxing robots and stuff like, come on, what are you talking about?
Ron GlozmanIt doesn't even make sense. No, think one of the interesting things that I know I'm still I haven't formed my opinion, I think one thing that technology could enable is like a universal basic income. I don't know if that's good or bad. I haven't gathered enough data to make my decision yet, but I think definitely technology could get to the point where machines could provide for us and we could all have a rather good quality of life without having to do work. I think the interesting thing there is that should hopefully allow us to pursue more creative and artistic endeavors if you've always wanted to be a painter or let's say like I personally like photography.
I don't really have time to do photography and it doesn't really make me money. I don't take time off my job to do it because I still got to pay my bills. But you know, hypothetically, there's less work and I only work half the time or one quarter of the time and I still make a salary just as good, then I can spend the rest of my time doing my creative endeavors. Yeah, exactly. Yeah, I mean, I've got really strong feelings about UBI. It's the same old idea
James Benhamthat's been rehashed for the last two hundred years and it just never really works out well. And so I you know, because people you have moral hazard, right? Moral hazard is a really big topic that you gotta think about, and when you incentivize people to do nothing, it turns out a lot of people have to take you up on They're like, cool. I'm gonna do nothing. And then, you know, product productivity and output bottoms out. And so, you can go through We can go through 20 papers.
We can go through 200 years of history on this and it Yeah. But it's really not central to our argument today. What I am excited about is that your technology is changing lives and is helping people buy insurance. Good, good on you for this. We've talked about like, what's the next big release that you guys have coming that you're really excited about? Just to wrap it up. It's funny you say that. It's actually something we're gonna call quote compare. So in the same vein is like even upfront,
Ron Glozmanit's like which one of these is gonna be the best to meet my needs? Because oftentimes they'll offer a lower coverage than you're asking for in terms of a limit or the deductible might be higher and there's sort of these little gotchas that make it really hard to read insurance policies and on the personal line side we're probably all familiar with like you know GEICO and all of those different companies. But on the on the commercial side, there isn't really anything that does that. So it's a tool that we're currently working on.
We have an MVP of and we hope to bring to market in q one. You're tackle work comp? Yeah. That's actually the one of the main lines that we start with.
James BenhamAlright. Sure. Well, I expect a phone call later. We're gonna Alright. I've spent the last fifteen years in work comp, so we're gonna talk about work comp software all day long. I've been listening. I would love your input. Yeah. You have a wealth of knowledge. Yeah. Yeah. I've been been I've been building work comp software a long time. It's a it's a it's a fascinating you know, work comp is one of the one of the more challenging lines in general because you have, know, these indemnity payments and you have all this medical stuff, and you're like, you guys just have so much to do with medical, and it makes it really complicated.
It really does. You're dealing with people and when they get back to work, and it's it's really, really challenging. Very, very tight. You know? And, course, The United States really complicates it because we have a different regulatory system for every state. And, so you really have 50 countries here. And, you know, from an IT perspective, it you know, there's part of me that's thankful for it because it creates a lot of work for technology companies. But the other part of it Yep. I'd really like to fix the problem.
I I'd rather work on other stuff. Like, you know, you know what I mean? Like, I'd rather fix other problems other than having to constantly update 50 regulatory charts. You know? But it's but but it is what it is. I mean, there's there there's a lot of opportunity there to to automate things. And and quote comparison would be really cool to actually be able to side by side Mhmm. And catch because, you know, where you get really caught up in quote comparison is the exclusions. And and in in particular, like, if you were able to tackle cyber, and that and that would be one of the harder ones.
And I'm sure you've got a cyber policy, and I do too because we're tech companies. Right? Cyber gets really dicey because they often include e and o, and the exclusions get really, really finite minute gotchas. You know? You can get in big trouble in cyber with the exclusions and with work comp. So I'm eager to see how you handle it. That's actually what I'm saying. Yeah. I mean, let's
Ron Glozmanchat offline. I'm happy to set something up. I mean, it's definitely, and that's part of the complexity. And that's the interesting thing is, you know, I mentioned we've looked at depending on how granular there's a variety of different touch points to work on. We've known about quote comparison for three or four years, but the reason we haven't built it to date is part of it is the complexity that you just highlighted. So you can only imagine what that means the other 14 mean if we're doing this as the third one.
Yeah. Right? That's awesome.
James BenhamMan, I enjoyed our conversation. Yeah. It's been a pleasure. Yeah. Thank you for having me. We've had we've had a bunch of Israeli entrepreneurs on here. You're an Israeli Canadian entrepreneur, so I'm digging that. I'm digging this. It's like it's like the best of two great countries. I love Israel. I love Canada. I love the entrepreneurial spirit in both places. It's really, really exciting. I'm fired up for you and and the product that you're building, and obviously, you've raised a bunch of money, so congratulations on that.
And, you know, just hold on to your dream, keep chasing it down, and and certainly we'll we'll follow-up with you as you keep, blowing it up.
Ron GlozmanSounds good. Thank you so much. And I I will say, you know, I I'm a sole founder, but I have to give it to the team. It wouldn't be possible without, you know, the many people that support me. So I can't take all the credit for all those wonderful things you said. Just have to throw in,
James BenhamI have to throw in the team as well. Good, good leader credits his team, and I appreciate that. I I've got 230 amazing people behind me that just absolutely do a great job every day for us here at JBKnowledge, and I'm I'm thankful thoroughly thankful for them, and I'm thankful for your team too because they're helping improve the insurance industry. And so you and I get to talk on the on the on the video cameras, but they're the ones, you know, make making the magic happen.
So, a big kudos to your team. And, a a big thanks to you for joining us today. Thanks for all of our listeners out there in listener land for joining us. That's right. For the It's Your Tech Geek podcast powered by Jimmy Knowledge at jimmyknowledge.com. It's all about technology that's transforming and disrupting in a very good way the insurance world. I've been your host, James Benham. Big thanks to Jim Greenlee, our podcast producer, Cara Dalton, our creative producer, Adele Waldeck, transcriptionist. And thank you for joining us today.
We're taking you on a journey through insurance tech. So enjoy the ride and geek out.

