Using AI to enhance financial services: lessons from Australia’s Moula

A study conducted by Microsoft Asia and IDC Asia/Pacific released findings which found that that organizations with AI expect to see 41% improvement in competitiveness in three years.

The study also found that more than half (52%) of organisations within the region’s Financial Services Industry (FSI) have already started on their AI journeys. This is higher than the Asia-Pacific average of 41%, indicating that the sector is more advanced than others in the region.

“The digital economy has resulted in demands for organizations to reinvent themselves such that they remain relevant to their customers. To do so, FSI organizations need to address three key imperatives – how to leverage data and AI for their operations, how to build and maintain trust among their customers, and how to tap on partnerships to drive innovation to stay ahead of the game,” said Connie Leung, Senior Director, Financial Services Business Lead, Microsoft Asia.

AI leader Moula

One key example of an AI leader is Moula, an Australian founded organization that uses AI to assess business loan applications made online. Recognizing the importance of small and medium businesses to Australia’s economy – most of the country’s 2.3 million businesses are classed as SMB – the company established an Azure based real-time credit decisioning service and leveraged Azure AI and machine learning capabilities to predict the probability of the SME being able to pay back its loan. Successful applications can result in business loans of up to $500,000 being made available in 24 to 48 hours, whereas the process through a traditional bank takes around 10-14 days.

Unlocking the power of a company’s data

Perhaps more significant than the faster process time is that, according to Moula Chief Product and Strategy Officer Justin Green, for many businesses, traditional banks simply don’t work.

“If you don’t have the benefit of assets, like the family home to put up as collateral, then your ability to borrow money from the bank is pretty limited. So Moula’s thesis was very much ‘can we partner with small businesses?’, try and demystify this volatility that exists in small businesses, encourage this open sharing of business data, as the critical asset to understand the nature of someone’s business,” says Green.

What Moula recognized was the value of the data that small business held about itself. If there was a way for lenders to see that data, analyse it, and assess the opportunities and risk – in short, determine the creditworthiness of the business – then it would be possible to identify who to offer loans to, because the data revealed that the underlying business was a good bet.

Green noted that small business was growing ever more comfortable with the idea of transacting online and engaging with fresh sources of support. The actual process of applying for a Moula business loan takes about 7 to 10 minutes. It involves completing an application form and granting approval for Moula to access the applicant’s business data. That’s generally achieved through an API into a data source, such as Xero.

From there the data is accessed, the business is analyzed, the risks assessed, and a decision taken about the requested loan. “It’s typically a 24- to 48-hour underwriting approach, once we have all of the business data provided.” Green states.

AI and machine learning as tools for success

Moula started testing Microsoft’s Azure at the beginning of 2018 before transitioning to the Microsoft cloud in October that year. This collaboration sees Moula leveraging Microsoft’s technologies for analytics, innovation and security.

Moula worked with Microsoft partner, BizData to plan and execute the transition. Together they established an Azure based data warehouse using Azure SQL DB, implemented Power BI to run business analytics, and leverage Azure AI and machine learning capabilities to assess data and analyse risk.

According to BizData Director Nadav Rayman the solution helps to empower small businesses by allowing them to use their own data as evidence of the quality of their business. “Moula’s partnership with Xero streamlines access to that data while Microsoft Azure and the surrounding Azure services provide an AI driven platform for Moula to rapidly analyse and assess a business’s loan application.

“Microsoft Azure is also acting as the foundation for ongoing innovation at Moula – for example the opportunity to create finance solutions available to the small business at the point of scale with Moula Pay.”

Moula’s credit decisioning relies on a logic it has designed, applied to the data made available to it through the lending process. This includes the application form details as well as accounting, banking, tax and credit bureau feeds. This helps create the quality accurate inputs needed by Moula’s machine learning algorithms.

Well aware of the value of that data Moula deploys multiple layers of defense in order to ensure data is protected. As part of its security controls, Moula takes advantage of Azure security features such as Advanced Data Security and Dynamic Data Masking.

FSI development and changing regulations

Operating in a currently self-regulated part of the financial services market Moula nevertheless expects to be impacted by changing regulations affecting other institutions.

Pesavento, Moula’s Chief Data Officer, expects that the arrival of open banking in Australia will accelerate the interest in alternative financial solutions as people and business grow used to greater choice and flexibility.

“Whilst customers already have the ability to share their banking data with us, the move to open banking should make this practice more common place and trusted, and help customers feel more in control of their data and able to use that data to more easily and more quickly access services from non-bank providers including Moula.