Thank you, everybody. Appreciate it. My name is Dylan Becker. I am the research analyst here who covers the vertical software space, but within that, insurtech, connected operations.
Use the mic.
Oh, sorry. Insurtech and connected operations. We have Rohan and Jean-Noël here from Roadzen. Rohan's got a presentation that he's going to run through, and then we'll have a handful of minutes for Q&A at the end. With that, I guess all the necessary disclosures you can find at williamblair.com, the obligatory remarks there. With that out of the way, Rohan, you want to go ahead? Thanks.
Hi, everyone. My name is Rohan Malhotra. I'm the CEO and founder of Roadzen. It's a company I started about 10 years ago. I'll start with a little story. My background is I studied robotics and AI at Carnegie Mellon, graduated 2010. I did my master's there and started building AI products for Fortune 500 companies right out of college. I'd built a business that had U.S. and India components. I was traveling back to India in 2015, and my friend had an accident on the road, and I was in the car behind him, so I sent him in my car to the hospital. Before leaving, he told me, "Look, can you go into the glove box and just file a claim for me? The insurance papers are in there." I went into his glove box. His car was almost totaled.
It took me about four and a half hours of making calls to the claims department of the insurance company, to the tow truck provider, to the garage that would accept the car upon being towed. It was honestly one of the worst experiences, and it just felt to me, why isn't there any technology that's helping me in this situation? If I paid my premium, it's to help me during this exact situation. That idea essentially became the genesis of Roadzen. I would say there are a couple of core theses behind Roadzen. We do AI for auto insurance, and we think there are a few tailwinds here which make this a really exciting opportunity. Number one, that AI will transform almost all of knowledge work.
Insurance is the best kind of knowledge work because every decision is made based on what we call historic data patterns, and AI is exceptionally good at doing that. The second thing is that the car itself is going to go from largely being a mechanical product that takes you from A to B to becoming what is essentially like an iPhone on wheels, and you can get data out of it, you can get decision-making out of it, and that changes the paradigm for the entire industry. We think that auto insurance, which is roughly a trillion-dollar market, is going to go from legacy-first players. In the U.S., the average age of an auto insurer, a top 20 auto insurer, is 95 years, to digital-first, AI-first players like Roadzen, and we're here to give you the story of how we think we're going to achieve that.
In the insurance world, you've got to make four key decisions. Number one is, how do I price this policy? Number two is, how do I sell this policy? How do I process claims, Number three, when they arise? Number four, I want to do all this and make money. In the world of insurance, that's called combined ratio. It means if I collect $100 of premium, how much did I pay out in claims, in distribution costs, in marketing, in my own administrative costs of running the business? That's expressed as a %. In the U.S., average auto insurance is at 104% combined ratio, which means that they are losing $4 on every $100 of premium they collect. That has been the pattern over the last 10 years. How do they make that up?
They can reinvest the float and make some money on that. That's how they cover up the difference. What Roadzen is doing, we are making insurance a great operating business again by reducing combined ratios, by making the car safer on the road, by helping on claims. We've built a vertically integrated technology platform that's powered by over 300 AI models that intersect at various different parts of the value chain across underwriting, making the car safer, on the claim processing, on the administrative aspects of insurance, and that's what we are continuing to build on. A few highlights here. As I said, it's roughly a trillion-dollar market. Roadzen operates in the U.S., Europe, and India. U.S. is the largest market in the world, Europe is the most profitable, and India is the fastest-growing.
We think this is a great combination, and it will allow us to build a very large business in a massive market. Roadzen has also an award-winning AI research lab. We've been collecting data for the past 10 years. We have proprietary data, over 300 AI models that have been built by us and are now being used by some of the world's largest insurance companies, car companies, and fleets to make decisions. It's also very important to understand we're not an insurance company ourselves. We either act as a technology company selling our tech to insurers, to fleets, to car makers. We have zero B2C component to the business. We are an enterprise technology player, or we act as a broker going through B2B2C channels to reach the customer. We'll partner with fleets, we'll partner with OEMs to get the product across.
Finally, we have seen our revenues grow pretty well over the last five years, from about $1 million a year. Last 12 months, we have done $50 million of revenue, and we have clear visibility to achieve $100 million revenue run rate in this coming year. We'll explain all of this in the next part of the conversation. One of the other things I think super important to note is that while we act as a broker or a technology provider in the insurance vertical, we never carry any underwriting or balance sheet risk on our own balance sheet. We always partner with insurers or reinsurers who carry the risk. It's a business that is asset light, does not carry balance sheet risk, and we believe now that we've built the technology, can be scaled to a massive level.
Auto insurance is a mandated product in almost every country in the world. It's a $900 billion market, 1.5 billion vehicles on the road, as of 2024. This is growing 6%-7% a year. This is roughly going to be a trillion-dollar market. Roadzen is building the leading company at the intersection of what we call AI insurance and auto, and we want to be the largest players in this space. Today, we have 390 people across nine global offices. We are headquartered in San Francisco. We have offices in New York and San Diego in the U.S., London and Paris in Europe, and we have a pretty large technology and AI research team of data scientists, PhDs in India. Almost 100 people that are part of our lab that build these AI models that we work on the insurance vertical with. Here's exactly what we do.
If you look at insurance today, it's like the advertising industry in the late 1990s or early 1990s. Everything is done based on demographic profiling, which is that how much premium should you pay is based on your ZIP code, your gender, your age, and what's your credit score. That's 95% of your premiums comes from that. 14% of drivers cause 63% of accidents on the road. 86% of drivers are essentially just subsidizing those 14% drivers. What we are creating is a personalized, real-time, dynamic driver scoring methodology that gives you how you drive, where you drive, when you drive, and augmented data around underwriting. It creates a highly personalized risk pattern, that we think makes insurance way more profitable and allows our clients to have positive selection on their portfolio, which we think is the best thing to do as an insurance company.
Secondly, since we are a broker as well, we have licenses in 14 countries around the world. We are licensed in the U.S., in India, and pan-Europe. We can actually sell policies to any customer in the world through a single global technology platform, collect the premiums in any currency, and pay out a claim in any currency, all of it in real time. That allows the ability to scale the business globally. The third thing, we focus a lot of our energy on this. Insurers care about collecting more premiums. Customers care about just one thing, which is: what is going to be my claims experience? We've built the best claims technology stack in the world.
Going from, if you have an accident on the road, just as your phone can tell you how many steps you walked in a day, we can get the data from a car, from an app on your phone, or from cameras around the vehicle that allow us to say there's a potential accident here. Send you a push notification on your phone so you don't need to call someone. It's completely proactive. Push notification says, "Are you okay? Would you like to file a claim right now?" Asks you to take a 360-degree video of the car. As you're taking the video, our AI is acting in real time, looking at the full video and saying, "This part is damaged," and then making a repair or replace decision, which then creates a real-time estimate of repair on the vehicle.
This process itself could take weeks with an average insurer. We are doing it in under two minutes, we went beyond creating just the estimate. We now have an ERP for workshops, which gives us access to real-time data into what are the parts available at the workshop. Are their bays open so they can repair the vehicle? Are there enough mechanics available? We can direct the car to a workshop that has a slot available, mechanics available, and the parts available, and get it repaired with a guaranteed 48-hour turnaround. We are doing now 3 million claims annually without ever touching the car on our technology today, which is one of the significant claims tech pieces in the world. Finally, I think one of the key things we'd like to do is not just be reactive, which is what insurance companies are.
I sell a policy, I wait for something bad to happen. It's to be proactive and continuously in charge of the policy, which is we're using telematics to not just reactively define the claim process, but proactively engage the driver, give them alerts. We have a division which is called drivebuddyAI, which is essentially a dashcam with an Nvidia chip on it that allows us to, in real time, say that this driver may be falling asleep, is about to enter into a collision. We issue an alert, three seconds. A human driver needs 1.3 seconds before an accident to act. We are roughly tripling their reaction time, and we've proved over 4 billion miles of real-world driving data, a 72% reduction in the number of accidents using our technology.
This technology has some incredible growth catalysts right now, and we are scaling it on a global basis. One of the things I think Insurtech 1.0 was all about what we call multi-line distribution businesses, which is that I have Jean-Noël as a customer, I'm going to sell him health and life and travel insurance and auto and bundle all of it together. Roadzen, we don't do any of that. We focus on building the best technology company in auto and insurance and just focus on this one vertical, which is a trillion-dollar market. All of this is powered by our AI research lab. As I said, there are 100 people in our lab.
We are one of the founding members of the AI Alliance, and what we do in AI is very important because when you hear AI, everybody today thinks it's about large language models, but that's not what we do. We build small models. We call it enterprise intelligence models that do one thing with exceptionally high precision, 95%-99% precision, because that's the only way you can deploy it at a large insurance company or a large car company. The way we do that is that we're going to say this model is just for KYC of the customer. This model is just for claims estimation. This model is for anti-fraud detection during the claim process. Singular models, and we built over 300 models that essentially automate judgment and workflow inside a heavily legacy industry like insurance today.
We've processed now 4 billion miles, over 20 million claims. Roadzen is the founding member of the AI Alliance, alongside Uber, Meta, ServiceNow, NASA. We were one of the 20 founding labs in the AI Alliance. Roadzen was also Anthropic's launch partner, one of a select few under 10 companies who were given early access to their managed agents platform to build agentic interfaces or headless interfaces around underwriting and claims in the insurance vertical. We are continuing to expand that partnership now. We operate in almost all the large insurance markets globally. We are continuing to expand. We have licensing capability in all our markets, and we act as a tech player in all the markets that we operate in. Today, we have about 4,150 global clients, which includes 140 enterprise clients.
When I say enterprise clients, that means it's a large car company, a large insurer, or a fleet above 1,000 vehicles that is using Roadzen's technology. Some of these players you will recognize, like eight of the 10 largest car companies in the world. Mercedes, BMW, Toyota, Volkswagen, are clients of Roadzen. five of the 10 largest insurers, like AIG, Generali, AXA, Allianz, are clients of Roadzen. Fleets like Supergas, which is the largest LNG transportation fleet in the world, are client of ours. Amazon is one of our clients, we are continuing to build this client base on a global basis. For car companies, what we deliver is that car companies want to embed insurance products alongside the vehicle ownership experience.
Being a regulated broker with underwriting capital behind us, we can give them a licensed product that they can sell to their customers underwritten on our technology, then when there is a claim, it comes back to their Toyota workshop, which is what they want. That's where they make money. For insurers, we give them the capability to underwrite with precision, to process claims without fraud, and do it much, much faster. For fleets, we work with them. Fleets want everything. They want their fleet to be safer. They want lower insurance premiums. Commercial auto in the U.S. has gone up 7%-8% annually for the last 10 years. In the last four years, it's been the number one factor in inflation as determined by the U.S. Economic Bureau .
Auto insurance is becoming a real problem if you own 12 fleets or more, and we are working to reduce that cost for our partners. As I said, we have two revenue lines in the business. The first is we sell our technology to our enterprise clients, and we make a per-transaction revenue. You, as an insurer, if you use the tech to process a new policy, we'll get paid $10 a policy. You use us to process a claim, you pay us $150 for every time you process a claim. If you're a fleet owner, you have 25 vehicles in your fleet, you pay us $25 per vehicle per month in that fleet.
53% of our revenues are what we call technology or enterprise AI revenues that come through this transactional revenue, and all contracts of Roadzen are three to five-year contracts, which gives us pretty good visibility over the next quarter, over the next year, and over the next three years on our business. Similarly, we have the digital broker or an MGA. An MGA is somebody who has been authorized by the insurance company to underwrite on their behalf. Even though our balance sheet is not at risk, they're giving us capacity to go out and say, "Go and do $50 million of business. You can price the policy. You can pay out the claims on my behalf, and I'll audit you once a year." That's what an MGA is. We're an MGA in 14 countries around the world.
That gives us the ability to control this entire experience from selling a policy through to claims, and we earn commissions from our insurance partners, which are roughly between 15% of the premiums to 35% of the premiums as revenue. Those are the two revenue line items. 53% today comes from technology and 47% from commissions as a broker. We grew revenues quite substantially over the last several years. I'd like to talk about years 2024 and 2025. Our year ends in March, so when I say FY 2024, it's April to March 2024. The U.K. regulator, we were doing $47 million roughly revenue at the time. The U.K. regulator said that car companies, when they sell insurance, charge too much commission.
If you're buying a BMW for $100,000, your insurance could cost $1,000, but the dealerships and car companies add on their commission on top of it and sell it for $2,000. They asked those guys to pause selling all insurance for two years alongside the purchase of a car. That contributed roughly $29 out of $47 million in revenue for Roadzen at the time. Our clients were the two largest car companies in the U.K. We kind of got hit with a gut punch three months after going public. We actually, though it looks like the year was flat, we lost $29 out of $46 million in revenue, yet ended the year roughly flat. We have really picked up pace over the course of the last 12 months. Our last quarter ended December, was $14.5 million revenues with a $0.5 million EBITDA loss.
We are expanding the business, and we'll be break-even this quarter, and we are continuing to grow the business from this point on. We have said publicly that we will achieve over the next 12 months, going from $14 million-$15 million a quarter to about $25 million a quarter of revenue. That's about 60%-65% revenue growth, and profitability. We have great confidence in this projection because since April 1, we have announced $20 million in new revenue deals. In the U.K., we signed nine different contracts worth $2.5 million, one contract with one of the top 10 car makers in the world for a multimillion-dollar contract for their U.K. insurance business. In India, we won a claims mandate from a top five insurer to manage one of their regions' claims process. That contract is worth at least $10 million annually.
They have $800 million of claims. We charge roughly 6%-7% on a per claim basis, so just one regional contract is worth $10 million on that. In the U.S., we signed insurance capacity from a tier 1 insurer with an A-rated carrier of about $30 million. That should lead to about $6 million-$7 million of additional revenues. We are very confident if December was $15 million, March will be better than December, and since April, we have announced $20 million in new contracts. We are very confident about our bridge to this $100 million or $25 million per quarter revenue and profitability from this point onwards. We got this data. We got some help from the Blair team.
There are 90 companies in software land that are less than $2 billion in market cap, but there are just three companies that are growing above 60% a year. Roadzen is going to be one of them. We have an incredible team, a global team. Jean-Noël, who's our CFO, he comes from one of the biggest reinsurance brokers in the world. I run the overall business. We have U.S., U.K., and India heads and an incredible board that supports the business. We try to marry experience in insurance, in AI research, and automotive inside the company. With that, I'll leave you guys with a few catalysts for the business. We think the stock could materially re-rate at this point. Today, Roadzen trades at a market cap of $160 million or roughly 2x next 12 months revenue. Our drivebuddyAI system has been mandated across India and Europe.
The governments have said that every new commercial vehicle needs to have driver safety technology in them before they are sold starting in April 2027. In India, Roadzen is the only company that has qualified the road testing, which is when they take your tech out for a spin inside a new truck, and they do it for four days. They bring 20 different drivers. It has to clear real-world testing before they authorize you to sell this tech. We are the only company. There are 1 million new vehicles that will be sold in India next year, commercial vehicles that require this tech to be embedded. We charge $200 a vehicle. That's about a $200 million new revenue opportunity opening up for us next year. We did the same thing in Europe.
There are 7 million vehicles in Europe under a new law called EU 2144, which is coming up. That's about a $1.4 billion opportunity. Roadzen is the only company in the world that has qualified road testing in Europe and in India, two of the major geographies for this new regulation. We were included in the Russell Indexes. We're going to be added to the index at, I think it's June 26th when the rebalance happens. Our market makers tell us that's about 6 million-8 million shares of buying. We think that will improve liquidity of the stock, improve investor recognition, and should be generally a pretty positive catalyst for the company. I also want to mention one interesting twist here. A Nasdaq company trades at a value of $160 million.
We have a substantial India business, which is about 50% of our global business. We went to some of the best capital markets investors in India, Quant AMC, Valentis Advisors, Prime Securities, and they invested in our India subsidiary, which is half of our business, at a $280 million valuation. We still own 93% of India as Nasdaq holders. We believe the stock is fundamentally undervalued at this point. Finally, I think the idea is we believe this is a business which will go from $50 million-$100 million. There are very few AI companies in the world that have $100 million of revenue. Very few still are profitable at that scale, and very few of those have recurring revenues.
A lot of companies have 50 million+ of revenue in AI, but those are POCs because every company in the world is giving you a $5 million contract to try out AI. Roadzen's revenue is 100% recurring in nature and directly linked to outcomes on the business. With that, I'd like to thank you all for listening to our presentation and look forward to any questions. Thank you.
Perfect. Thank you, Rohan. I know we have room in the breakout. I think we've got a couple of minutes, so maybe one or two questions to start. Very helpful context. I guess if you could give us a sense, we walked through the market dynamics that are impacting insurers and the automotive ecosystem today, but that network platform differentiation of your vertical integration, right? You're the only one who can see where the assets are, how the operator's behaving, what that means from a risk perspective and underwriting, and how that actually gets solved from a claims perspective. Maybe dive into the differentiation of that data set and why it's not something that's easily replicable or something where you see other competitors going to try to do a similar function, if you will.
When we look at the competitive landscape, there are people who just do underwriting products for insurance. There are people like CCCIS who just do the claims. Nobody has married all the data that the fleet safety feeds into the underwriting, then the claims data feeds back in and creates a holistic picture, and you control the whole outcome. This has taken us 10 years to develop over billions of data points on the full outcomes of insurance. While you may say that there are competitors to Roadzen, there are telematics companies out there are underwriting AI companies out there are claims management companies out there. Roadzen is very unique that we see the entire data from the entire value chain, and it's the only way that you bring the combined ratio down. There's no other way to do it.
If you underwrite something and you don't know what the claims are doing, you're half blind, right? You really need to see the full value chain. This is why we say that if the average auto insurer in the U.S. is at 104% combined ratio, we have been operating in the sub 93% for the last six years. We are 10 percentage points better than an average insurer, and in many cases, significantly better than that.
That's probably something that unlocks incremental investment and market share gains from those carriers that are utilizing your platform. I guess maybe for Rohan or Jean-Noël here, just to bring you in as well too, very impressive financial profile and inflection that you've called to. The two segments of the business, I guess how you think about the mix shift happening or changing between the enterprise software side and the MGA brokerage side, and maybe the unit economics of who you're selling into with the majority of the business serving India today, but how that moves into North America, the customer profile and what you're seeing there.
About 50% revenues for us come from India. About 30% is now the U.S. U.S. will become our biggest market in about 18 months. We are seeing tremendous growth here. The macroeconomics are the most interesting thing from our perspective in this, because in India, an average auto insurance policy, there are more two-wheelers than four-wheelers. The average auto insurance policy costs $65. In the U.S., it's like $2,000, right? In India, we get paid $16-$18 a claim. In the U.S., we get paid $150 a claim for doing the exact same thing on our technology. Not only is the U.S. the biggest market, but it's also the most profitable market, and we are seeing tremendous margin expansion, profitability expansion, as well as top-line expansion as we are continuing to grow in the U.S. market.
India, which is growing 70%-100% a year, but as we continue to build out the U.S. market, we think in 18 months' time, that becomes the biggest market or over 50% of our revenues.
Fantastic. Sir?
I was going to agree.
Perfect. Well, with that, I believe we are at time. We will continue the conversation for those that have additional questions upstairs. Rohan, Jean-Noël, thank you very much.
Thank you.