Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Reuven Shnaps on Outstanding Claims with the Price Writer Ep 10
37 Plays1 year ago
Reuven Shnaps from Earnix is our guest today on Outstanding Claims. We hear how he originally aimed for academia before consulting and then becoming an integral part of the success story of Earnix. The conversation touches on:
🎓His journey from a PhD at the University of Pennsylvania to the insurance industry.
💼Insights into the early days at Earnix and the growth of a small startup into a global enterprise.
🎯The importance of considering, not just risk premiums, but the whole pricing strategy.
Watch now to hear about Reuven's amazing start-up journey.
Recommended
Transcript
Introduction to Earnix and Its Value Proposition
00:00:00
Speaker
Every time I was able to convince people that what we do will bring them value, why they need to license or purchase a solution like Earnix, that was for me a success.
Meet Reven Schnapps: Career and Role at Earnix
00:00:11
Speaker
Today I'm joined by Reven Schnapps, the Chief Analytics Officer from Earnix. So let's meet Reven. Hello Reven, welcome to the show. Hello Jeremy, glad to be here today. Thank you for having me.
00:00:24
Speaker
Brilliant to have you on as well. Reven, could you tell us, how did you get where you are today? Yeah, it has been a long journey. So maybe I would say it starts when I was around 24, decided to go to the University of Pennsylvania to pursue my PhD in economics. With the plan that I will go back and be a service as a university professor at the department of economics, Barreland University.
Earnix's Growth and Industry Leadership
00:00:51
Speaker
plans alone, and I guess the reality was different. I ended up spending about four years working at Anderson and Deloitte and Touche. That was most of my time. And then about 18 years ago, I met Sami Krikler before we were planning to go back to Israel. And what I would say, the rest is history. I had a unique opportunity to join the company. It was then a small startup.
Advancements in Pricing Solutions
00:01:20
Speaker
and see it grow over the years to become one of the leaders of a pricing solution, pricing rating solution for the insurance industry and also Earnixes, you might know, also operate on the banking side. So it's really been a very interesting journey.
00:01:39
Speaker
What's that been like, starting with a company small and actually seeing it grow to be a big business? Yeah, that's, I guess if I go back then, no one would even have an eye that was the wrong numbers all my life and kind of dealt with predictions and stuff like that. I would never know that I would be in the company 18 years later, but really I think it's been an amazing journey. And when you start up
00:02:09
Speaker
At the year of 2005, things were very different. The economic situation was very different. The challenges were different. And I think along the years, we had a vision, we had kind of a plan to come and contribute and change things for the insurance company. And I would say one of the things
Real-Time Pricing and Technological Adaptations
00:02:29
Speaker
that was unique
00:02:30
Speaker
You probably know that you come from that background is that there was a focus on. I would say a cost plus approach, right? People were on the cost side with earnings. I think broke Sam is a semi-curricular. The founder ideas was to say, Hey, when you think about pricing, there are these two sides to it. You have, of course, the cost side. You have to have kind of the revenue side. You have to understand how customers and view.
00:02:57
Speaker
pricing, how you have to take into consideration competition. So on top of the discipline that you had for insurance pricing, estimating or building loss cost models and stuff like that, came the side of the customer behavior models. How do customer respond to rate changes? Because you might change your rates as a result of risk.
00:03:22
Speaker
but customers know nothing about your risk models. They never left the home and you are telling them they are risky and all they see is now that you have changed the rates. So this is something you need to account for when you want to put market prices.
00:03:39
Speaker
So I think that was very unique aspect. So focus was initially on more of what you can call the pricing sophistication. Companies have different business objectives and what's the best way
00:03:53
Speaker
And what kind of set of prices would you put in the market in order to be able to achieve your business objectives? Of course, you had some people focusing on profit, some were focusing on growth, and you had to be able to balance between the two. Of course, over the years, we also learned that pricing is a very complex kind of process. A lot of stakeholders.
00:04:17
Speaker
And it's not just about the sophistication, it's also about bringing those into the market in real time. So, Earnix is a rating solution. So, how do I take all those great analytics, kind of, and all the sophistication and make sure that it hits the market in the right time, that people get the rates in real time and there is different requirements for that. So, over the years, Earnix has built
00:04:44
Speaker
not only the sophistication, but also the processes to be able to support it. Of course, many years ago, it was around moving to what we call an online system that in a matter of milliseconds provide a quote, then later come also the cloud solutions. So you're starting to build the infrastructure in order to be able to support very large scale of operations. And Ernest today works with
00:05:14
Speaker
a lot, numerous insurance companies across 35 countries and if in the over six continents and you have the insurance tax, you have the startups and also you have the mega insurance companies, the top 10 carriers in the US sample.
Defining Success with Earnix Solutions
00:05:32
Speaker
What's been the biggest success for you with earnings over the years?
00:05:38
Speaker
That's a very good question. There was a lot of talking about me personally, my own personal success. So I would say, look, so I would say that it's interesting. We are now in twenty twenty three in twenty twenty one. The markets were focusing on growth and earnings also.
00:05:56
Speaker
became a unicorn. So you could say that, hey, look, from a small company that declared Ionex as a unicorn, that was really a major achievement. But I will say, you know what, you don't need to look at the very big success every time.
00:06:13
Speaker
I was able to convince, I also participated in trying to convince people to adopt Earnix. Every time I was able to convince people that what we do will bring them value, why they need to license or purchase a solution like Earnix, that was for me a success. And when we actually went and worked on projects and showed and deployed it and showed them the value that they are getting from it, it was a success. Every time we were able to link
00:06:42
Speaker
what I call this analytics in order to address and solve a business problem. That for me was a success and I've had a lot of stories of this along the year. It's brilliant. What's been the biggest challenge?
Challenges in Insurance Industry Adoption
00:06:58
Speaker
What's been the biggest difficulty?
00:07:02
Speaker
I think the biggest difficulty, and I think we still see it, is that we are dealing with a fairly conservative industry. They do want to move forward. They do have great aspirations. I think over the last few years, they were challenged also by some smaller, at least their incumbents, by smaller issue techs and the digital revolution, corona and stuff like that. But still, it's a very traditional
00:07:32
Speaker
slow moving to some degree industry. And also, I think the other challenge is you have legacy systems that kind of delay the adoption.
00:07:45
Speaker
I think in along the years I also come across the people use that the data is not ready exactly it's a challenge you need to make business decision you need to adopt different methodologies and stuff like that and say data was a big kind of I would say not inhibitor but slow down the process along the the years but I do see a change if you compare
00:08:10
Speaker
Well, the industry was when I started in 2005 and maybe 2010, 15, even 2020 before Corona and today with the big advancement in the digital revolution, I would say, and ML, AI, and definitely now generative AI. I think people understand they didn't need to move much faster. They are going to stay behind.
00:08:36
Speaker
And that's where earnings come and to help them, to support their journey, to speed up the innovation, speed up the journey and some of the need to do some catch up. So that's, I think, my view on the challenges. That's great. And what would you say is your mission for insurance pricing?
Revolutionizing Insurance Pricing with Earnix
00:09:02
Speaker
Yeah, so I think when I think about the mission is that we, as I said, we have been a leader in this space for many years, and we want to revolutionize the way companies are operating. Now, operations, of course, you're talking here today about pricing, but it's not. It's pricing, it's rating, and later you can talk about underwriting processes and other aspects. And also,
00:09:29
Speaker
And especially in the last few years, it's not just about the operations, but how they serve the customers. Clearly, today more than ever, we see customers' needs is to have a personalized treatment. You need to tailor the treatment to each customer. And so that's one side there. The second is how do you respond to ever-changing market conditions?
00:09:56
Speaker
So I think when we look at those two, this is where Ernie's kind of mission is to really help them to, as I said, improve significantly the operation side, respond to customer needs in a personalized manner, and also be able to adapt to market conditions. Now, take, so one clear example was the Corona period, change everything we knew about
00:10:22
Speaker
what's happening. So you have to change, you have to change your models, you have to change your processes you need. If you don't have the right software solution and technology in place, you have a big problem. If we look at the last year, I think inflation is a very big concern. Now, in so many countries, so we have actually customers who are operating in a almost double digit inflation a month. Okay.
00:10:53
Speaker
and wow some places i think they described in south america you have and in of course in the in europe and the us it's not that high but even if you have a double digit inflation number that's something that they haven't seen also the cost of some of the parts is really dozens of percent so that means
00:11:16
Speaker
You need to be able to change your rates much more frequently, but you have to do it smartly. Yeah. Yeah. You can just go ahead and change sometimes called base rates or whatever. Just take, let's take 10% up and that's it. So you have to be very smart how you do it. That's why you need a technology that will help you make those changes quickly. But as I said, smartly, intelligently.
00:11:45
Speaker
Now, the other thing that is very important, and this is where this operational efficiency, we mentioned that this is a highly regulated industry, some places very strict, some place less strict, either way. So being doing this in a fully governed and controlled manner, it's extremely important. So what
00:12:07
Speaker
We do a focus also is how do you do all that, as I said, in a controlled and double manner and actually affect two things. One is how do I cut significantly my time to decisioning? So the ability to really take into account all the changes, all the update, the models and everything, that's one aspect. I need to be ready with the prices. But then the other one is how do I cut significantly the time to market?
00:12:36
Speaker
That's where I bridge. I have legacy system. I have IT prioritization that prevents me sometimes from bringing this to the market in a very quick manner. And that's where earnings also excel in both of those areas. And I think that's a noble, I would say, mission that we have for the industry. And another thing I think it's very important
00:13:01
Speaker
So since this is a complex pricing process, many stakeholders, so we need to have a mission to serve as a center, this common language, that all relevant stakeholders will find the solution useful and they can have an exchange and speak to each other through the Linux and channel the information and make sure we connect the IT
00:13:26
Speaker
all the way to the data analyst, to the pricing actuary, to the salesperson marketing, all of them can find this within the same solution. And another thing that is critical in insurance, and that's also part of the mission, we need also to do all that, we need to do it in a seamless way, and we need to minimize errors. That's very important. And one last point, I used to say when
00:13:57
Speaker
When I participated personally on a lot of pre-sales activities, you buy into earnings, you buy into the future. We need to make sure that when you are starting the journey with earnings, you can always know that we are keep innovating. And we, no matter where you are on the adoption journey, you'll always have the ability
00:14:20
Speaker
to be ahead because we are innovating and you can actually take more and more of the new things that we are bringing to the market. And one last point that is very important, it's not just a software play. You need to complement the great product with proven field experience. Ernix has been in business for over 20 years. I have myself 18 and there are other folks
00:14:45
Speaker
within the company so this you need deep experience and understanding of insurance how you bridge between business and analytics and we also complement those services by partnering with in a system integrators or consultants that really can together give a holistic approach to pricing and make sure companies are successful.
00:15:10
Speaker
That's brilliant, Reavan. I agree with all the points. I particularly picking up on deployment and making sure things are deployed correctly. It's so important and people don't like to admit mistakes, but it does happen. I know from personal experience, using earnings is good for preventing those problems from happening. What would you say is your vision for the future of insurance pricing?
Future Vision: AI and Machine Learning in Business Decisions
00:15:37
Speaker
I think what we see in the past few years, and definitely I would say in the last year, it's becoming even stronger, that companies realize that they need to leverage data and advance, make more informed business decisions and elegant decisions. We have seen ML and AI being used a lot around automation, process automation,
00:16:04
Speaker
But it's time to really take this to the next level. And we want to make sure that your business decisions are being informed by that. And in that respect, we see different types of, I would say we want to talk about before about personalization or tailoring it to different customers. We also have different personas. We have insurance companies that like to bring their own models.
00:16:31
Speaker
with the advances that you have out there. People wants to use machine learning more than the classical GLMs are not enough for them. That's perfect. Earnix is in that respect. Embrace open analytics and help them basically operationalize those models. And I think that's very important because most of the ML AI projects, I think Gartner has a statistic like 85% actually are failing. Yeah.
00:16:58
Speaker
Not because the models are not great. People are doing a great job. It's because they don't know how to operationalize them. They don't know how to deploy them. And we put a lot of emphasis on the one hand to give you the ability to really take the best models, but to make sure that they operate in scale and to make sure that they will be used in the market. So that's the one side, open analytics. The other side of that is that some people want to also accelerate adoption of ML and AI and they need assistance.
00:17:27
Speaker
There was a shortage of MLAI folks, so we can help them with built for purpose models like the auto GLMs or hybrid model that helped them to combine the great features of machine learning for accuracy, but at the same time taking the good stuff from the classical statistical model when it comes to inferencing and we can combine them together. So I think that's part of our vision. Another important when I think about
00:17:58
Speaker
leveraging ML and AI, we need to remember, as I said, we have different personas. And it's not always, or in many cases, the business people are the one who makes the decisions. And they need to be aligned. My vision of democratization of AI, it's a matter of trust. How do we help them trust all those great models so they can leverage those for decision making?
00:18:27
Speaker
And for that, we also invest myself and my team and the others within earnings in developing those solutions that can bridge between the great work that is being done by the analytics data scientists, actuary folks and the business people. So you have explainability solutions. The other part is that often you find that people are very focused on prediction.
00:18:54
Speaker
I want the best models. But when it comes to pricing, there is much more than that. You have to do the separation between prediction and inferencing. So we also provide tools for the business people to have kind of to diagnose the great models and see whether they are suitable for pricing. So at the end of the day, what we want to make sure is that everyone can bring the models they want.
00:19:20
Speaker
But we want to focus also on the deployment side. We want to make sure that they adhere to the principles of pricing and economics, if you like, and then can be used at the end of the day by business people to make business decisions and ultimately serve customers better. Another
00:19:42
Speaker
important, I would say, two other points. One is around generative AI. I cannot speak about this. Everyone talks about generative AI, LLM applications, and there is a lot of them. I'm not going to, we don't have enough time, I think, to develop this, but one area that we have started to work on is around this generation of synthetic data. I think data, we want more data,
00:20:08
Speaker
But there is a lot of concern about privacy and how you use it and how to share it externally. And I think this advancement around the generative AI when it comes to tabular data, the ability to really generate as much as you need data,
00:20:28
Speaker
is really opens up significant opportunities for insurance companies both for the smaller startups or insurance that may not necessarily have a lot of data available to them but also ability to really to share the data with third parties for research purposes or some even if you have large organizations they may not be comfortable sharing data internally. Also the
00:20:55
Speaker
ability to generate synthetic data and combine it with the real data can actually help improve the stability and performance of some model, especially when you have rare outcomes. So you need to complement the data. So really, I see an area that we are working on. And according to Gartner, I've seen kind of unbelievable statistics that in 2021, they said only 1%
00:21:23
Speaker
of data was used was synthetic, they predicted by 2024 around 60% of that will be, but I think it's really telling us that this is booming. And I would say finally people talk about dynamic pricing. Now I know this is a topic that has been talked about, but what does this really mean?
00:21:45
Speaker
So for some people, if they used to do pricing every six months and they've done now doing it every three months or two months or one month, that could consider the dynamic. I'm talking about the ability where you have first competition, the market dynamics, the changing environment. Is there really the ability to move to a real time kind of intelligent pricing? And one of the areas is true. You need intelligent dynamic monitoring system.
00:22:13
Speaker
that will catch arrows when on the fly, will alert where you have either the drift when it comes to the mix of data or where your models are not performing and be able to decompose those effects. And very important, this is for business. Remember, I have this in mind. I want to democratize AI and bring it to the business people. They need laser sharp root cause analysis and insight
00:22:42
Speaker
into what is happening. What's the problem? They want to be aware. They want to have a bit of mind when they go on vacation that someone is watching the back. Now, the real vision, how do I close the loop? How do I move from insight into providing corrective action? Maybe initially some suggestions, but for some models, when you have more trust,
00:23:07
Speaker
and maybe it's not mission critical and you have some regulatory requirements, you can actually have an auto pilot mode. Online learning of model being updated and business and pricing strategies also being updated automatically. That's a vision that I see for insurance pricing.
00:23:30
Speaker
That is excellent Reuben, thank you so much. It chimes so well with the how price rights have used this and my methods particularly that modelling is important but there's so much else that we need to consider as well to do this right.
00:23:46
Speaker
Thank you so much for joining me this afternoon. It's been really great to hear from you. Is there anything else you'd like to add? No, I think we have covered a lot of ground. I think it's an exciting time when you have ups and downs. It's actually where opportunities arise. And I think those will have the right technology.
00:24:07
Speaker
and business strategy I think are going to prevail. We are here of course to support the journey of companies and I would say don't wait for tomorrow, start the journey today. Absolutely and the price strategy is so key to that as well. Thank you very much Reivan, you have a good rest of your day. Okay thank you, bye bye.