Computer Vision in Fintech

If you are familiar with computer vision solutions, you are probably surprised by the title of this article. What has computer vision to do with finances and the fintech industry? After all, these are two completely different things, aren’t they? As it turns out, not necessarily! Computer vision in fintech is currently a dynamically developing field of knowledge, and every year we see more and more applications of this technology. Actually, even now, computer vision can significantly improve the way the modern financial world works. Today, we will show you how your company can utilize this tool in order to improve your offer and customer service.

Let’s start with the beginning. We talked quite a lot about computer vision technology in the article about Computer Vision in e-commerce. We encourage you to read it. Basically, computer vision is all about teaching machines, algorithms, and applications to process and recognize what is displayed on a screen or in an image. In other words, computer vision is designed to interpret visual data with no human assistance.

Further reading about Machine Learning in finance

How can it be useful for a fintech company? In this article, we are going to tackle four major applications you should be interested in:

• KYC verification
• New payment methods and banking data security
• Investment strategies
• Property analysis for insurance purposes

Let’s start with the first application.

Computer vision applications:

Computer vision enhances KYC verification

This abbreviation stands for Know Your Customer. KYC process comprises all the procedures designed to verify the identity, suitability, and risks involved with maintaining a business relationship with a given person or entity[1].

Companies utilize KYC processes for the purposes of ensuring their potential customers, agents, consultants, or distributors are anti-bribery compliant and are actually who they claim to be. If you run a financial company, you fully understand how vital KYC verification is.

Unfortunately, today, this procedure is quite time-consuming. At best, it takes several hours to verify one person/entity. What can computer vision do here? It can significantly accelerate and automate this process! We want to show you two real-life examples. Similar solutions can be implemented in your company!

The first example comes from BBVA. This bank allows customers to open an account using a smartphone. During a video call, a customer verification process takes place. It’s a quick and efficient method.

A bit different, but still a similar approach has Revolut. As you probably know, it’s a modern online fintech with no physical branches. If you want to open a Revolut account, you have to take two pictures–one of your ID and one of your face. Next, Revolut’s computer vision algorithms verify these two images and compare them. If everything seems fine–your account is open and ready to use. If not–you are asked to take additional action in order to confirm your application and identify yourself.

Most likely, you already see the fantastic benefits of this solution. Because everything happens with no human assistance, human employees of BBVA or Revolut can concentrate on other, possibly more complex or urgent tasks. It’s a convenient solution for the customer as well! Because computer vision algorithms work 24/7, the account can be opened even in the middle of the night, and it takes just several minutes to finish the entire process! It’s a clear win-win situation for your company.

New payment methods and banking data security

When it comes to payments, customers have two essential requirements. They have to be quick and safe. And this is what computer vision offers. Let’s think about modern payment solutions. Usually, you don’t have to carry a credit card anymore. You can pay for your purchases using your smartphone and virtual credit cards (again, it’s a solution Revolut offers).

Now, this technology can be enhanced even more! Computer vision has everything it takes to help banks remove the need for credit or debit cards. More and more financial institutions worldwide start to offer the possibility to withdraw money from the ATM without the need to input a credit card. Again, all you need is a smartphone! But this begs the question, what can happen if someone steals your smartphone? First of all, you have to take care of the safety of your data.

Companies like Apple, Google, and Samsung have taken one step forward. Their devices can be secured by biometrics verification. It’s a verification method based on the physiological characteristics of a given person (yes, that’s possible thanks to computer vision). For instance, let’s take the Samsung Galaxy Note9[2] smartphone.

The available biometric authentication features are fingerprint scanning, face recognition, and iris scanning. Similar solutions are available in other modern devices. Thanks to this solution, banking data stored on a device (for example, in the banking app) is 100% safe–no one other than you can access your phone.

Banks can utilize similar solutions, and, for instance, introduce biometrics verification in their mobile banking apps, to make sure that only an authorized person has access to money and data.

Investment strategies enhanced by the satellite image analysis

Satellite images contain vast amounts of data, especially when it comes to investment purposes. If a company wants to buy a site for their new commercial or residential building, satellite pictures can tell a lot about the surroundings, and current state of this site. All of that without the need to go there.

Or maybe another example; let’s say a large commercial network wants to build a new store in a given city. The consumer count for the already existing stores is a useable knowledge! With computer vision and satellite imagery, companies can assess that, based on the number of vehicles in the parking lots. If you can analyze how many cars park in other competitive stores in this area, it’s much easier to assess the investment profitability. As a result, your company can reduce the risk of planned investment.

That’s true when we talk about just several pictures, human specialists can do that with ease, but what happens when you have hundreds or thousands of such images? You need AI to speed up the entire process. And this is precisely what can be done thanks to computer vision algorithms. These algorithms can use geospatial imagery from satellites, UAVs, and airplanes to obtain priceless insight into market trends and investment opportunities at the global and regional scale.

This is what Orbital Insight[3] does. It’s a geospatial data analytics company. Orbital Insight takes satellite data sources and helps their clients obtain useful business insights using image processing, machine learning algorithms, and cloud computing.

They offer geospatial intelligence, supply chain monitoring, real estate due diligence, and other services to make more informed decisions about planned or current investments. They claim that they can look at historical satellite imagery and, based on that, draw useful conclusions about economic trends in the given state or country.

Property analysis for the insurance purposes

Computer vision technologies enable much more accurate underwriting, taking into account more aspects and elements, like the exact real estate specification, state of the building, and other features. Again, satellite imagery comes in handy. It allows insurers to view all the crucial property attributes at the time of underwriting. With computer vision, insurance companies can immediately validate property features, which significantly accelerates the quotation process, mainly because it removes the need for physical inspections. Furthermore, insurers can continuously monitor their clients’ properties and see if any changes are applied during the period of insurance.

Cape Analytics is a property intelligence company, based in Mountain View, USA. This company makes it accessible for insurers to diligently verify the state of each property. They provide insight into granular details about properties that materially affect risk, valuation, loss, and probability of any problems. In order to do so, they use structured data extracted from high-resolution imagery using machine learning. Moreover, to make the whole process even more accurate and straightforward, they have a ready-made database of 110 million properties across the US.

All of that means huge savings for your company! We can expect that the need for human real estate inspectors will be significantly reduced, and insurers will be able to send appraisals much more quickly. We will gladly help you implement property analysis algorithms in your company!

Major benefits and challenges related to computer vision in fintech

As always, this technology is not flawless. Naturally, there are some significant benefits and advantages of using computer vision in the fintech industry, but there is also the other side of the coin. It’s time to sum up this article. Without a shadow of a doubt, computer vision can help eliminate many problems that torment modern financial institutions, perhaps your as well:

• Tons of paperwork
• Lack of time

Computer vision is quicker and more efficient than other traditional methods. It helps financial companies automate a multitude of processes. Computer vision also helps to protect personal and financial data and protect money against theft.

Finally, computer vision can be used for property valuation or insurance claims verification. All of that can be done quickly and correctly. Utilizing computer vision in your company means huge time savings and the improvement of many processes.

Are there any downsides? Unfortunately, yes, but they are not caused by the technology per se. The very first one is strictly related to customers’ habits and expectations. Many of them, especially older people, are skeptical or even distrustful of this technology. They are simply used to more traditional methods. Therefore, banking companies have to pay a lot of attention to customers’ education, which consumes time and money.

Next, banks should invest in high-end cybersecurity solutions to ensure that all of that sensitive customer data is safe and resistant to possible outside attacks. And finally, adopting such sophisticated technology is associated with large investments and reconstruction of the IT infrastructure. Nonetheless, it’s the future, and sooner or later, computer vision solutions will be commonly used in the fintech industry. In your company too, we have no doubt about that!

If you are interested in computer vision, get in touch with us. We will gladly show you all of this technology’s aspects and help you implement it in your enterprise. We are at your service!

References

[1] Wikipedia.org. Know your customer. Apr 16, 2021. URL: https://en.wikipedia.org/wiki/Know_your_customer. Accessed Jun 24, 2020.
[2] Samsung.com. What are the biometric authentication features on Galaxy S9, Galaxy S9+ and Galaxy Note9?. URL: https://www.samsung.com/global/galaxy/what-is/biometric-authentication/. Accessed Jun 24, 2020.
[3] Orbitalinsight.com. Global Insights. Anywhere. Anytime. URL: https://orbitalinsight.com/ . Accessed Jun 24, 2020.