Jakarta. A Singapore-based company has enjoyed being the sole player in the debt collection business that employs artificial intelligence in Indonesia, and perhaps also in Southeast Asia.

AsiaCollect arrived in Indonesia in November 2018 and has since attracted many big clients including the country's top banks, its executive told the Jakarta Globe in a recent interview.

More and more Indonesian companies who do not have enough time and resources to deal with their non-performing loans have turned to AsiaCollect for help.

"We are providing a digitalized debt collection. There are so many fintech companies doing online lending: peer-to-peer, payday loan and many other types. They all want to lend, but most of them don't know how to collect or get the money repaid. We're helping them to do just that," AsiaCollect's country manager, Guillermo Martin, said.

The company uses artificial intelligence to get to know borrowers better and avoid using coercive means in collecting debts, he said.

The debt collection business in Indonesia has gained notoriety for employing threats and even humiliating the borrowers. Recent media reports said loan recipients in online lending schemes were often subjected to public humiliation when they failed to repay. Lenders could easily do this since they require loan recipients to provide their social media usernames and passwords before their loans were approved.

But Martin stressed his company would never do such things.

"If others want to do it, that's their choice. What we want to do is reputable and transparent debt collection. We want to help financial institutions recover their money, but to do it in a way that is not damaging to the customers and doesn't have a negative impact on society," he said.

"We are using technology as a tool to increase efficiency. We have developed artificial intelligence and machine learning to let us know more about each person with non-performing loans," Martin said.

AsiaCollect's operators use speech recognition technology to create automatic psychological profiles. The system will recognize the language the borrower uses, the tone of the language and how frequently he or she uses certain expressions. 

Based on the profile, the operators will learn if the borrower is psychologically "positive" or "aggressive," Martin said.  

"What we are doing is emotional recognition... recognizing what your emotional state is like at that very moment. The operators use this to put you in your comfort zone," the Spaniard said.

"We have automated quality control so the machine can detect bad language and threatening language," he said. 

Martin said AsiaCollect has introduced a more humane approach in collecting debts by educating borrowers who sometimes have no clues about their financial state.

"Sometimes we speak with them and they ask us: 'How come I borrowed Rp 1 million and now I owe Rp 4 million?' They don't understand their financial predicament, so [sometimes] we are not just collecting but also selling a repayment plan. But at the same time, we are educating them by saying 'Look, this is the interest rate, this is the penalty. If you don't repay in this term, then you get a penalty.' That's our approach to debt collection," Martin said.

Top Clients

AsiaCollect Indonesia employs 45 operators – all locals – and its services are available across the country although they never leave their office at the Anugrah Building in Mega Kuningan, South Jakarta's elite business center.

"We can cover all of Indonesia from our call center here. We have people working 8-to-8 in two shifts of eight hours each so we can operate in Indonesia's three time zones," Martin said.

"We have the capacity to employ up to 130 operators. We are still hiring because more and more companies are coming to us for help. At the moment, we have 16 clients. Seven out of the top 10 banks in Indonesia are working with us," he said. 

AsiaCollect's clients also include two of Indonesia's largest multi-finance companies, five major fintech companies and two e-commerce companies.

"We are trying to be ethical and reputable and we only want to work with companies that have the same approach," he said.

Martin argued it was bad lenders who had made the debt collection business so notorious in Indonesia. 

"If they are unethical lenders, they are going to be unethical collectors. These are the guys who are threatening people, harassing people and calling twenty times even in the middle of the night," he said. 

"It's bad for me because it creates the image that debt collection is a bad activity when it doesn't need to be," Martin said.

Not Traditional

AsiaCollect's AI-based system allows its operators to remain at the office while calling borrowers. The company does not believe that face-to-face collection is effective. 

"We don't do field collection because financially it doesn't make sense. We don't believe it works. We don't believe people should be threatened [to pay debts]. I don't think I need to go into your house to embarrass you in front of your neighbors," he said. 

The company has two ways to make profits. One, it can partner with lenders who give their debt portfolios and pay collection fees. Two, it can buy non-performing loans and collect them for itself. In Indonesia, AsiaCollect focuses on the first business model, Martin said.

AsiaCollect now operates in Vietnam, Indonesia and India. Singapore serves as its headquarters only since the fintech sector in the city-state is too small.

The company also plans to start operations in the Philippines and Thailand next year.

"We want to stay in Southeast Asia. We have very powerful investors, some of them from Thailand. So they're always calling us, 'Hey, when are you going to Thailand to help us out here?' So we'll most likely expand to the Philippines or Thailand or both," Martin said.