More and more companies nowadays simplify and accelerate business processes through some sort of automation. In a 2020 McKinsey Global Survey, for example, two-thirds of respondents said their organizations were at least piloting the automation of business processes in one or more business units or functions, up from 57% two years earlier.
Artificial intelligence (AI) and machine learning provide further fuel for market advances in automation. These technologies enable solutions to automate a growing number of business processes, from instant payment initiations to identity verification checks.
In this blog, we’ll explore how you can jump on the bandwagon and start automating 5 bank card-related business processes via our APIs and SDKs.
Which cards can be scanned with OCR and AI?
Before we turn our attention to the business processes that can be automated, it’s good to know which bank cards can be scanned. Well, the answer to that question is relatively simple.
Our card scanning solution works basically on all sorts of bank cards containing printed or embossed texts, numbers, and logos. These elements can be read and data can be extracted from it automatically. A smart card scanning solution is therefore extremely handy in taking over the manual process of reading and checking bank cards, and the related data entry, usually performed by clerks.
The most commonly used bank cards are credit cards and debit cards.
Credit cards
Credit cards are convenient for everyday purchases and have the advantage that they are widely accepted for purchase, for example when you travel internationally. They allow you to borrow money from the card issuer up to a certain limit in order to purchase items or withdraw cash.
From credit cards, we can extract the following data fields:
- Card type
- Credit card processing network
- Credit card issuer
- Card holder name
- Card number or PAN
- Expiry date
- CVC (optional)
- Signature (beta)
Debit cards
Debit cards are generally easier to obtain and set you on a budget, meaning you cannot spend more money than you have. They allow you to spend money by drawing on funds you have deposited at the bank.
From debit cards, we can extract the following data fields:
- Card type
- Debit card network
- Bank name
- Card holder name
- Bank account number or IBAN
- Card number
- Expiry date
- Signature (beta)
How does automated credit card scanning work?
Now that we know which cards can be scanned, let’s have a look at how our solution works. It might become a bit technical from here, but we will try to explain it in layman’s terms.
- The first step in the process is providing a picture or PDF file of a debit or credit card to our API. This can be done with the mobile app by simply taking a picture with your smartphone. We will automatically remove the background and correct any rotations. It is as easy as that!
If you do not have camera functionality in your app yet, the Klippa Camera SDK might be beneficial to help with bank card scanning. This is a full-camera scanning kit that has square recognition, cropping, perspective correction and quality detection.
- As soon as the picture or PDF file has been received, the image is optimized and converted into a raw text file. In this step, all the text from the card is extracted, but it is not structured yet. You can see an example below:
- Next, the Klippa Parser takes the text file from step 2 and converts it into a structured JSON format. The JSON is then returned as output from the API.
From here, the extracted data can be used according to your own needs, such as passing it on to your database for further processing.
All of this is completed fully automatically in just a few seconds, without any human intervention. That sounds pretty nice, right?
Why should you use automated credit card scanning?
Next to being fully automated without any need for manual actions, our card scanning solution has many practical benefits that can help your business move forward. We will list the most important ones below:
- Reduce costs
- Increase speed
- Reduce errors
- Prevent fraud
Reduce costs
Manually checking and processing bank cards can be quite a labor-intensive process, both for you and your customers. Just think about the manual data entry, card verification, and archiving requirements that come with it.
Our solution can save you a lot of time, and thus labor costs, in this process by automatically extracting data from bank cards. There’s no magic formula for this, of course, but intelligent automation typically results in cost savings of 40% to 75%, with the payback period ranging from several months to several years.
Increase speed
It only makes sense that automated processes are faster than manual processes. Take a process like credit card verification. When done manually, reviewers have to go through several documents, make sure the information is correct and check for fraud. Humans aren’t anywhere close to being as fast as computers and automating this process can save you a lot of time and money.
Reduce errors
Did you know that manual data entry inevitably leads to data entry errors? Error rates typically range from 0.55% to 3.6%, although error rates as high as 26.9% have been found. Typing on a smartphone isn’t the most enjoyable experience either, so these rates may be even higher for mobile devices.
Luckily, automated processes and computer programs do this substantially better. With our credit card scanning solution, you are assured of the best data extraction software for bank cards, thereby minimizing the risk of human error.
Prevent fraud
Our card scanning solution not only includes OCR technology but also contains AI-based image tampering detection. This involves detecting incoherent pixel structures, such as strange changes in lighting or sudden color changes in text. As a result, the software detects forged images in seconds, making it easier to filter out fraudulent applications before they can cause you any trouble.
In addition, our system can cross-check the information on a person’s card with external resources. If you have blacklists or fraud databases with known fraudulent cards, for example, these can be consulted automatically.
Common use cases of automated debit & credit card scanning
We try to extract as much data as possible from debit and credit cards via OCR and machine learning. This ensures that we can support as many use cases as possible. Below you will find the most common ones:
- Simplify credit card payments
- Digital customer onboarding with credit cards
- Automated age verification via credit card scanning
- Bank account (IBAN) validation through debit card scanning
- Automatic anonymization of bank card data
Simplify credit card payments
On many websites and applications users have to type in their credit card numbers for purchases. This can be quite an annoying task and a typo is easily made. Imagine how easy it would be if users only had to scan their credit card with their phone’s camera!
Capturing credit card information is as simple as that with our AI-powered credit card scanner. Users breeze through payment forms by just taking a picture or uploading a PDF file of their credit card to our API. With the right OCR and data extraction models, fields are identified, localized and the text in them is automatically extracted and entered into payment forms.
The UX of an application or checkout will dramatically improve, because users don’t have to enter the information themselves anymore. With our credit card scanner, form-filling will be easier than ever, leading to higher success rates of the checkout and eventually higher revenues.
Digital customer onboarding with credit cards
In order to ensure a stable and compliant service, many online payment and trading platforms require a person’s credit card information at sign up. A primary concern here is verifying that the person trading on the platform is actually who he says he is and that he is using a valid, authorized credit card.
While many merchants still use a manual process to perform these checks, a growing number sees the need for automation and operational efficiencies. OCR, machine learning and AI are considered the best technologies to accomplish these.
In the cryptocurrency industry, for example, the identity of customers is verified automatically by validating the authenticity of their credit cards and matching the information on them with their identity documents (e.g. name of the card holder).
Automated age verification via credit card scanning
As part of the customer onboarding process, companies may carry out identity checks to ensure they onboard the right customers for age-restricted services, such as alcohol or tobacco websites and video games with objectionable content. This can be done by requiring users to provide their credit card information and verifying the data.
Credit card verification requires credit card holders to prove that they are in possession of the card using methods such as Card Verification Value (CVV), Address Verification System (AVS), and 3D Secure. This binds ownership to the person buying online, with the knowledge that the customer has to be over the age of 18 to apply for a credit card.
Although this depends on the assumption that the vast majority of credit card holders are adults, it’s definitely a more sophisticated age verification system than simply asking users to fill in their date of birth on a form.
Bank account (IBAN) validation through debit card scanning
International Bank Account Numbers (IBANs) are widely used in domestic as well as international payments. Being a fairly long string of digits and numbers, the IBAN is often a point of error when manually entered into forms.
With our software, companies can validate that the IBAN actually exists and identify individual bank accounts. It helps to reduce failed transactions due to mistyped IBANs by identifying data structures and performing format, digit, and length validations.
Companies can thus verify an IBAN and even provide proof of IBAN ownership by matching the cardholder’s name on the bank card with a person’s identity document.
Automatic anonymization of bank card data
Over the years, companies may have collected thousands or even millions of documents containing privacy-sensitive information, such as bank card data. More often than not, they are not fully aware of what is actually in there, let alone being able to navigate through them effectively.
With the help of Klippa’s OCR API, these companies can automatically detect specific combinations of numbers in documents, such as IBANs or credit card numbers, and have them anonymized by removing or blacklining those specific lines on a document or deleting the documents entirely.
We can also anonymize data right from the start by automatically blacklining certain elements after a bank card scan, such as the last 3 digits of the IBAN. Normally, this is a tedious, expensive and burdensome task, but with our smart solution, it becomes as easy as 1-2-3.
As such, we help businesses to speed up their bank card-related processes, stay compliant and provide a hassle-free experience to their users.
What are your options when looking for a card scanning solution?
Your first thought might be to build the entire solution from the ground up. Regardless of your in-house team of developers, building such a solution yourself demands time for research and testing, funds for your crew and office space, and most of all, a change of focus from what your company normally does.
So it is rather expensive and complicated to opt for this option. Why do it yourself when you have proven solutions available for direct implementation?
The Klippa bank card scanner, for example, is available via an API and SDK (Software Development Kit). That means you can easily integrate our scanner into your existing software in order to make use of its functionalities.
The following part is probably just for developers, but to give you an idea of the JSON response that you receive from our API, we show a simplified example below:
And that’s not all. Our API is very easy to implement. To prove our point, you can find a PHP code snippet below. Of course, the API can be implemented in different programming languages as well.
We have documented our API carefully and provide both business and technical support if needed.
How can Klippa help in automating your business processes?
We hope that you are convinced of the value of a card scanning solution after reading this blog, but you still might have some questions. Is Klippa the right solution for your business? Can we help you with your unique use case? How difficult is it to implement and start using our software? And so on.
Our experienced product specialists are more than happy to answer all your questions. They can tell you everything about what our solution can do for your company and how you can integrate it seamlessly into your existing systems and operations.
You can reach out to us directly or you can plan a free 30-minute demo below, in which we show you how our software works and how it will benefit your organization.