Throughout the pandemic, new and used car sales have surged, causing a shortage in inventory but how do you know what your car is worth? Is it time to trade in and buy or lease something else? carVertical, the blockchain car history registry can help. Rokas Medonis, CEO of carVertical shares more details with us in this exclusive interview with TechBullion.
Please tell us your name and a little more about yourself?
My name is Rokas Medonis. I am 35 years old and currently am a CEO and co-founder of IT data company carVertical.
After high school, I started my career path as a photojournalist. In 2009 my colleague and I created and launched a photo news agency in the Baltics. We took our market share quite fast. This was a turning point when I decided to stop being a newsboy (I was a very decent photographer) and shift to IT projects. Job after job, and I became CMO at the big dating site. In 2015 I decided to go into max risks and try to start-up businesses by myself.
Many failures later, I’ve ended as CEO at carVertical, an IT data company that creates products from historical car data. We are scaling rapidly. At the moment, almost 80 people are working at the company. We are covering 25 markets and expecting 16mil EUR revenue in 2021.
What is carVertical and How does a blockchain car history registry really work?
Fundamentally people tend to cheat. They managed to cheat in all epochs in all possible fields.
When a car loses it’s warranty that the manufacturer provided, it starts losing data on itself. As a result, dark spots of its history tend to appear. And that’s where unfair car resellers come into play. They use those dark sports to leverage the price to their side by doing some dirty works – clocking odometers, hiding damages, or doing other dirty jobs. In other words, used car sellers and resellers are often cheating.
carVertical technology gathers all possible data on used cars and tries to cover all those dark sports. Also, we are implementing machine learning on our data to analyze it. Our goal is to beat the cheating seller out, protect used vehicle buyers from fraud, and provide them with a bargaining chip.
We are putting all gathered data on a particular car and putting everything into a data report that any client of ours can use and read.
I will show how it looks in reality by using our sample report: https://www.carvertical.com/sample-report
How has the pandemic affected new and used car sales, how is carVertical helping your customers meet these challenges?
In the first year of pandemics (2020), new and used car markets boomed. Everyone understood that the rules of the world had been changed instantly. Moreover, people understood that social distancing is the new reality. That means if you live in a big city or metropolitan area – forget about public transport and get yourself a car.
But in 2021 everything shifted. Worlds logistic problems and the crisis of semiconductors made a big impact on manufacturers. As a result, deliveries of new car orders are delaying and that leads to people not selling their current cars. It strongly affects the second-hand car market.
carVertical sees that the rate of unfair practices in the used cars market has increased significantly. So our products in this industry is a life-changer. And we see that reflected in our sales figures.
How has blockchain helped carVertical develop the most advanced search engine for used cars?
Blockchain is not an industry-changing technology. It is expensive, hard to implement and adapt. Nevertheless, it was and still is a buzz keyword. Many industries have tried blockchain and burned their fingers.
However, for our data business, blockchain fits very well. Of course, we are still looking for deeper implementations, but the main benefit is data and authenticity validation.
carVertical has become the leading product in the historical car reports business and a well-known brand in just as little as four years. And obviously, problems come with that. Unfair car sellers started cheating and counterfeiting our reports.
At the moment, we are utilizing blockchain as an authenticity validator. For example, we are putting blockchain records for every report that we are selling and proving that there was generated report for a particular car.
What is a VIN decoder?
A VIN decoder is an online tool that extracts specific information about a particular vehicle. There are plenty of details encrypted in the 17-character long alphanumeric string, starting with the assembly plant the car was manufactured in, ending up with original equipment the car is equipped with or even characteristics of that car:
- Engine type
- Engine displacement
- Power output
- Transmission type
- Drivetrain specifications, etc.
However, we can clearly express it is not that simple to decode the VIN number since cars for different regions or manufactured in different assembly plants have different VIN structures.
Luckily, every car has a unique VIN number and that allows us to track the history of each particular car. Many years ago VIN numbers were used just to separate unique vehicles, but now people can access much more details about that particular car just with a VIN alone.
How does carVertical use Convolutional Neural Networks for object detection and image classification to extract essential vehicle information from images.
The Convolutional Neural Networks (CNN) are mainly used to transform an image by using filters. Let’s imagine that we have a picture of a building. By applying different filters we could highlight the edges or the shapes of the building. The different representation of the same image provides the chance for the CNN model to extract important features from the image. These features help the model to distinguish a building from the rest of the pictures. The same applies to the vehicle images as well.
carVertical is using an extensive database of vehicle images to train CNNs to:
- – detect license plates and blur them;
- – identify a person in a picture so we could protect individuals identity;
- – identify different types of vehicles: cars, motorcycles, trucks, etc.;
- – stipulate vehicles condition: damaged or not;
- – extract car features: make, model, body type, etc.;
- – detect important vehicle information: documents, VIN sticker, insurance, etc.
All these Deep Learning models, prepared by carVertical Data Scientists, helps us to extract multiple vehicle features, protect personal information and increase the quality of the reports.
Tell us more about your projects on Blockchain/DLT solutions, what should we lookout for?
Tokenomics for data collection
Vehicle data has value only in large quantities. If you have data for a single vehicle, we can only pay pennies, which is insignificant to a user. To collect more unique data directly from users, we extensively use tokenomics. Using a combination of economic incentive models with our token, we provide enough benefits to incentivize even single vehicle owners to collect data about their vehicles. Unique data, in turn, help us to achieve our mission – fight fraud and protect used car buyers.
Tokenomics for data validation
User entered data must be manually validated by people. We are training machine models to automate this, but we still need people. We are employing tokenomics to rewards users for data validation. It uses a combination of economic systems to achieve this.
We are building a solution using blockchain to verify data authenticity. It used to be economically unfeasible because of the amount of data. But, the advancements in Blockchain Layer 2 solutions allow solving this issue without breaking the bank while using public blockchains like Ethereum.
We are building more and more products where users record their vehicle data. We are building an ecosystem where users can share and sell consent to their data. Using public blockchain is the only correct way of doing it transparently.
There is a problem in European Union that all countries have different standards for vehicle registration and transfer while the cross-country vehicle market in the EU is massive. It is difficult enough to transfer a vehicle from one person to another in a single country. It is even harder to transfer cars between people from different countries. We are working in collaboration with multiple EU countries to achieve transparent vehicle history by unifying data standards. After we standardize data, we will standardize vehicle registration and transfer standards to allow borderless vehicle transfers possible using a public blockchain.
Where is the car data sector moving on?
It’s tough to predict what challenges are ahead of us. On the biggest is connected cars trend. Car manufacturers tend to gather as much data straight from cars (without permission of car owners) and keep it to themselves. Also, it is an opportunity because when someone tries to monopolize all the data, Robin Hood type of businesses (such as ours) comes into play. We have open opportunities for strategic investors to help us meet our targets for this market.