Artificial intelligence

What is artificial intelligence?

The digitization in the world in general and the development of lightning-fast computers has over the past 10 years helped pave the way for artificial intelligence (AI). But what is AI really?

AI is an area of computer science that works to develop intelligent machines and software that think and act humanlike. Voice recognition, learning, perspective, planning, and problem-solving are some of the areas that scientists are trying to emulate.

You must distinguish between two different types of AI: strong and weak. Weak AI is characterized as AI-technology that can deliver limited human actions. An example of this is Amazon’s voice software, Alexa, which can change songs on Spotify or adjust the light in your room. Strong AI, on the other hand, is software that thinks and not just simulate human intelligence. So far, strong AI is just a hypothesis as it has not been invented yet and we do not know if it will ever become a reality.


The interest is growing!

You may not notice it, but AI is already in use commercially today and most likely to a greater extent than you think. The technology is used in a variety of areas such as image recognition, drug development, search engines, e-commerce, self-driving cars, and more. Artificial intelligence also helps to improve the customer experience on several of the online platforms that you use daily. If, for example, you are a user of the music service Spotify, you may have experienced that you are suggested new songs that suit your taste in music.

Apple’s speech recognition software SIRI is also an example of AI technology used by many people worldwide. SIRI allows you to ask questions and enable different actions by voice demands such as sending messages, solving math, making phone calls, etc. We are already taking advantage of the technology, whether we think about it or not.

AI technology has many areas of application and the potential behind the technology is huge. AI is probably one of the technologies that are going to affect business and our daily lives the most in the future. Both Danish and international companies have noticed the many application possibilities, which to a greater extent focus on artificial intelligence, machine learning, and the work with big data. Collectia is no exception in this context, as we believe that technology can benefit our debtors and clients in many areas.

Collectia vs. traditional debt collection agencies

In traditional debt collection agencies, debt collection cases often follow a static and predetermined collection process that is believed to work the best. This method does not consider the type of debtor or the circumstances of the case. This often constitutes a long and inefficient recovery process, which often results in poorer recovery. But at Collectia, we do things differently.

In the video, you can learn more about how artificial intelligence and dynamic recovery processes enable us to make the right decisions at the right times in the recovery process.

Effective debt collection process – every time!

Meet our data expert, Cai

We would like to introduce you to our colleague Cai (Collectia Artificial Intelligence) – she is a girl who loves data and loves to learn new things. Cai’s role in Collectia is to analyze and process data and knowledge that we collect about the Danes in our IT system.

Cai analyzes vast amounts of data, and she does so lightning fast. Based on the data collected, Cai can decide how we should put together the most optimal recovery process and at the same time choose the form of communication that the individual debtor prefers. Some prefer to receive formal letters, while others prefer digital communication. For example, we see that young individuals prefer text messages that contain payment links that make it easy and quick to get their debt out of the world.

With the combination of Cai’s knowledge, our debt collection system, and the latest digital payment solutions, such as Mobilepay, SMSPay, and our user-friendly self-service solution for your customers, we have enhanced debt recovery to a level not seen by traditional debt collection agencies.

The advantage for you as a customer is that we can recover your unpaid invoices faster, better, and more professionally, as we hit your customer on the media and forms of communication that your customer prefers. Furthermore, Cai can recommend if a case should go to legal proceedings or not. Here, Cai can calculate the outcome of a legal pursuit.

AI and machine learning are not just something we talk about at Collectia. It’s something we use daily. Cai is getting smarter day by day, and so are we – to the benefit of both our clients and debtors.

Unique collection score

Our BI department (business intelligence) has developed a unique debt collection score, which indicates debtors’ ability to pay on a scale from 1-10. The score is based on more than 2.5 million debt collection cases that we have processed over time. The ability to pay is calculated by looking at the Debtors gender, age, payment history, payment remarks in RKI and Debitorlisten, case type, demographic, and 25 other parameters, which ensure a statistically valid assessment. A score of 1 suggests that there is a high probability of payment. A score of 10 suggests there is a very low probability that the person is willing to pay the debt.

We believe that all people are different and should be treated individually. The score is a powerful tool that our debt collection expert, Cai, uses to create our individual dynamic and situational recovery processes, which ensures an effective debt recovery. The individual collection score can change from day today. Therefore, we update and feed the score with new information daily so that we can ensure that we use the right recovery activities that are most likely to trigger a payment.

Is debt collection driven by AI in the future?

The work with artificial intelligence across industries is a continuous process that over time will result in the creation of increasingly more advanced IT solutions. The debt collection industry is no exception. Here, debt collection agencies are in search of technologies that can provide better customer experiences, streamline the customer journey, and improve the overall debt collection while taking the needs of the individual debtor into account.

Initially, several companies in the industry focused on optimizing the recovery process by analyzing big data and implementing champion/challenger programs to identify the most effective recovery processes based on debtor and case type. Now the focus has shifted to the development and utilization of the technology behind chatbots, including virtual agents, to streamline case management in the recovery process as well as to automate the communication flow.


What is a chatbot and why are they important?

A chatbot is software powered by artificial intelligence that can simulate conversations (via text or voice recognition) with users through chat programs, websites, mobile apps, etc. Chatbots are often described as one of the most advanced and promising forms of human-machine interaction. From a technical standpoint, however, chatbot simply represents the evolution of software capable of detecting and answering predefined questions. Some are very sophisticated, while others simply scan for keywords, and based on this, the chatbot delivers a response that matches the registered keywords or matching phrases.

A virtual debt collection agent

One of the newer inventions in the AI universe is virtual debt collection agents (a combination of AI and chatbots), which can communicate with debtors and perform case-related tasks.

Examples of tasks performed by a virtual debt collection agent:

  • Setting up installment plans
  • Present case information
  • Refer debtors to case handlers
  • Answering general questions

Now, virtual agents are more or less just an idea, but it is far from unlikely that we will encounter them soon. Collectia and several other financial companies are already looking at the possibility of realizing this dream.

Are virtual debt collection agents the future?

We believe that the next step in the work with artificial intelligence in the debt recovery industry is virtual agents who, with human characteristics, can act as case managers and negotiate repayment individually. The virtual agent of the future will be a further development of the existing chatbots that can be found in apps and on websites today. It will be able to simulate human emotions as well as characteristics and play an important role in the recovery process.

It is impossible to predict what the future will bring. However, one thing is for sure. Digitization, regardless of industry, will drive innovation and the competitive advantages of companies going forward. One of the major benefits of a digital recovery strategy and the use of artificial intelligence is the ability to leverage automation to determine actions in the customer journey (recovery process) where a real human interaction should be performed to provide the best and most efficient service. Although a virtual agent theory can handle large and important parts of the recovery process and communication with the debtor, there are situations where human intuition, understanding and decision making are necessary to streamline recovery and customer management.

Chatbots, including virtual agents, play a role in the overall automation of call centers and outbound departments of debt collection agencies, as they can combine machine learning with expertise based on the experience of real debt collection agents. Chatbots’ takeover of certain functions in the recovery process can potentially result in the following benefits:

  • Debtors get 24/7 access to a support function that can answer questions and help the user in the process.
  • Debtors can define preferred communication channels via the virtual agent early in the recovery process. Information that will subsequently be included in the dynamic recovery process to improve the recovery rate.
  • Debtors can talk about their debt issues in a safer environment, as they do not talk to a real person (our experience shows that some debtors feel uncomfortable talking about their debt issues).
  • The debt collection agency can automatically obtain information about the debtor’s behavior and the circumstances of the case based on indicators registered in the dialogue with the bot. This means that you can identify and initiate situation-specific activities such as proposing a call from a real collection agent.
  • The virtual agent can identify situations where human interaction is needed. The agent can then present contact information to the company support center.

Chatbots and virtual agents, we believe, will represent valuable tools in the recovery process to combine digital and traditional communication channels.