Interview Lula

How do you perceive the overall transformation that Generative AI can bring to the insurance industry? In what ways do you anticipate Generative AI enhancing the overall customer experience within the insurance sector?

There are numerous areas where AI will have a profound impact, but upon closer examination, it becomes evident that industries heavily reliant on human labor will undergo the most significant transformations. Despite considerable advancements in technology over the past few years, such as sophisticated underwriting processes, we've observed a troubling trend: an increase of 20% in loss ratios. In the US, the current combined ratio stands at 104%, with 65 points attributed to losses, indicating that 40% directly affects human labor.

This signals a critical conversation within our industry: How can we enhance loss ratios without resorting to massive layoffs? Is there a possibility that we've overstaffed, with perhaps unnecessary redundancies among our team of adjusters and investigators? This is precisely why we believe GenAI will play a pivotal role, particularly in streamlining repetitive and mundane tasks, such as determining minimum liability requirements or verifying information during the initial loss notification process.

It's essential to emphasize that AI isn't intended to replace human involvement entirely. Instead, it offers a solution to address the inefficiencies resulting from overstaffing. By leveraging GenAI for repetitive tasks, insurers can optimize their operations significantly. This marks a pivotal moment where we anticipate witnessing the most substantial impact and utilization of GenAI.

In our ongoing study, we explore three significant waves of technological transformation reshaping the insurance industry. The first wave witnessed the initial impact of digitalization with the advent of smartphones. The second wave, driven by real-time data and the fragmentation of the value chain, ushered in the era of Connected Insurance. Now, we are experiencing the profound influence of GenAI. How do insurers strategically position themselves amidst these simultaneous waves of technological evolution?

Insurers need to recognize that they aren't equipped to spearhead this transformation alone. In today's competitive landscape, top-notch Gen AI engineers are in high demand, and they're unlikely to gravitate towards traditional insurance companies. Instead of resisting innovation and attempting to develop cutting-edge technology internally, incumbents would be wise to embrace specialized insurtech firms focused on AI solutions tailored to the insurance industry. By doing so, they can tap into a wealth of expertise and insights from both technological and insurance realms, essential in navigating the intricacies of a highly regulated industry.

Over the past decade, we've witnessed a strained relationship between insurers and insurtechs, characterized by a misguided belief among incumbents that technology development is straightforward and a corresponding assumption among insurtechs that insurance is an easily solvable puzzle. This dynamic has created unnecessary tension. To move the industry forward, insurers must recognize the value of collaborating with insurtechs, while insurtechs should actively involve insurers in their initiatives.

Encouragingly, there's been a noticeable shift in insurers' behavior. While some still harbor ambitions of in-house development, many are now open to collaboration, signaling a growing understanding of the benefits of partnering with insurtechs.

Your company has been involved in developing customized AI solutions in Insurance. In your opinion, what are the best practices you have identified for seamless AI implementation in the insurance industry and what are the main challenges that Lula has encountered in the implementation process?

The biggest challenge is optics, AI leading firing friends. There's a widespread concern that AI will lead to job losses, especially in a sector like insurance where we've historically over-hired. However, AI can actually help make insurance more cost-effective by reducing the need for excessive staffing. With AI trying to make insurance cheaper, minimizing the people you have to hire, it ends up increasing profitability for your company and making it affordable for your clients, so the biggest challenge is that people think we’re trying to build something to replace them.

The second is seamless AI implementation. You have to build this technology in a way where it can be set up in 30 seconds by anybody. We have an AI that we call Gail (Gen AI by Lula), designed for an easy set up, and the very first case we developed is a conversational AI that can sell and service to your customers. And when people look at that they compare us with the other companies that are doing conversational AI use cases.  One of the things that catches my attention is the other conversational AIs have six months implementation times that probably implies a cost of 200.000 and 300.000 dollars. 

So the main reason for our success stems from our approach. We’ve developed Gail with the end user in mind, focusing on simplicity and user-friendliness. We created our product on October 26, we've funded through revenue, and our streamlined onboarding process has attracted more customers in three months than all other companies.

GAIL is a voice powered AI tool that is capable of having human-like conversations and has been considered as the first AI capable of passing the Insurance Licensing Exam. How do you envision Gen AI’s role in insurance sales, and what impact do you expect it to have on customer service?

We believe this technology isn't about replacing humans but enhancing their capabilities. Take brokers or agents, for instance. Traditionally, they spend around 10 hours a week qualifying leads, another 10 closing deals, and another 10 on outreach. With AI, lead qualification and outreach can now be automated, saving them 20 hours of work.

This means they can now devote 20-30 hours to higher ROI tasks like closing deals and providing better service to their clients. For example, if a sales agent has to make 1000 calls a week to fill their pipeline, they may not have time to answer urgent calls from clients needing to file a claim.

With our technology, like Gail, we're maximizing the efficiency of sales agents by freeing up their time from repetitive tasks. This allows them to focus more on their clients, providing better support during critical moments and ultimately improving their overall productivity and availability.

How has the advanced use of Tech Titans data by these companies influenced the overall data landscape in the insurance industry, and what lessons can insurers learn from their data strategies?

If you think about Tesla, Amazon and Google what is really impressive about them is not how much data they have but it’s how easy they make getting information from data accessible. For instance, in Google if you have a million pictures in Google photos and you want to find a picture of a horse you just need to type in horse and its gonna get all the pictures of the horse. Getting access to my information super easily is something that is not happening on the insurance side. If you ask a claim’s adjuster right now any of the big groups they suppose to have all your claims data, they’re paying millions of dollars in claims for X, Y and Z products, but if I ask any of those claims send me over the report, it’s going them over a month to put that report together and send it over to you. What insurance the industry has to learn is to figure out a way to make all this data somewhat usable in a way that gives us high quality information because right now since none of the systems connect, it’s useless and that’s why we have more data than ever but we see loss ratios continue to worsen year over year.

We’re doing something on the trucking side of things. If you’re underwriting a trucking company historically, you only underwrite that trucking company when renewal comes up. So once a year. We essentially made it per use, charging per mile basis or per use basis. For doing this we’re tracking their FMCSA numbers, which is a number assigned to a trucking company from the federal government, we’re tracking their vehicles, their drivers. SO if the driver got a DUI three months into the policy we can adjust the rate. That’s an example on how we use all this real-time data, constantly feed us different risk scores on people and if all of a sudden we have systems pinging this risk score or having surpassed a certain threshold then it was time to look at the price and reevaluate if we’re charging them adequately or not.

Tech titans play a pivotal role in shaping the AI landscape. How does Lula differentiate itself in the market, especially considering the influence of companies like Google and Microsoft, and what lessons have been learned from strategic collaborations?

Our main differentiation is being focused on Financial Services, specifically in Insurance. One of the interesting things about building AI within a very heavily regulated industry is that every other industry you can build AI to have 99% accuracy. If you are in Insurance you have to build an AI to have 100% accuracy and that’s how we really differentiate ourselves.

Tech Titans are probably more interested in diversifying across industries rather than focusing on just one industry, the way we do in insurance. They want to be the group that supports all types of industries, even if that means supporting the specialists that go deep into one particular one. This support strategy within all the industries like Amazon with AWS probably means getting a little piece of everything ends up with more money in their pocket in the long run and less risk.

Tech titans are increasingly forging collaborations with insurtech and traditional insurance players. How can such collaborations benefit the insurance industry, and what opportunities exist for insurers and insurtechs to leverage the technological expertise and data capabilities of these tech giants?

Let’s look at what chat gpt has done. Chat gpt has given the incredible technology that you feed it information and it writes it almost instantly. If you think about an insurance policy, even if you’re given the template, a new policy takes typically 4 to 6 months because it’s hundreds of pages, you have to go state by state and you’re typically looking at about 25.000 to 30.000 words, imagine the cost associated would be to have a team of lawyers, underwriters and insurance people spend four to six months building a policy for a new program from scratch. Now think of a tool like chat gpt that can write 25.000 words in less than a minute for 67 cents. So now you can build a policy from scratch within a couple of minutes and it may not be perfect, you might still need a couple of people to manage this, but you just reduced the time to build a new insurance product from months to a couple of days. And those are massive savings that are impacting everybody and those types of savings would not be possible if companies like Open AI hadn’t come out and made their tools available for such an affordable price.

That’s an incredible example of the way that Tech Titans are collaborating with not just Insurance and Insurtechs but companies and members of a society as a whole. 

With the rise of AI and data-driven decision-making, ethical considerations become paramount.

How does Lula approach the ethical use of AI in insurance, especially when it comes to customer data and privacy?

Our approach emphasizes the responsible and cautious use of AI. Insurance agents, as salespeople, are naturally incentivized to sell as much product as possible, particularly when their income depends largely on commissions. This incentive-driven behavior can significantly influence the type of insurance products customers ultimately purchase.

With AI, however, when it's trained with specific rules, it operates without bias or ulterior motives. Unlike humans, AI won't deceive customers to sell higher-priced premiums for personal gain. From a service standpoint, AI offers an opportunity to ensure buyers receive fair and appropriate insurance products.

In the US, statistics reveal that in some states, fewer than 55% of insurance agents pass their continuous education exams. This is concerning given the highly regulated nature of the industry and the importance of the products they sell. Additionally, 42% of Americans have had to make lifestyle changes due to unaffordable insurance. By leveraging AI to make insurance more accessible and affordable, we can promote greater responsibility in the industry.

It's crucial to encourage thoughtful consideration of AI, especially considering its novelty and ongoing development. Rather than hindering progress, we should embrace AI as a tool for improving accessibility, affordability, and responsibility in the insurance sector.