The Aboitiz Group, a family-owned conglomerate in the Philippines with interests extending from infrastructure to commodities to manufacturing, is now deploying artificial intelligence in financial applications.
David Hardoon, managing director of Aboitiz Data Innovation (ADI), says finance is the “torchbearer” for rolling out AI-backed services.
The group’s interest includes Union Bank of the Philippines (UnionBank), which also holds one of six digital-banking licenses in the country, which it operates as UnionDigital. Hardoon also serves as UnionDigital’s chief data officer.
Early uses of data-backed services include auto loans and bancassurance, but Hardoon says the same principles will underpin corporate finance.
Data career
Based in Singapore, Hardoon has built a career in data science. He has advised Singapore’s Central Provident Fund and the government’s anti-corruption law enforcement agency, as well as served as chief data officer at Monetary Authority of Singapore. He has a background as an academic and a consultant.
He joined ADI when it was created late last year. “I don’t want to talk about AI,” he told DigFin. “I want to do it.”
One of the opportunities for a large conglomerate in an emerging market is to use data and AI to extend the reach of financial services to the underbanked or unbanked, which he says includes around 40 million Filippinos.
To a data scientist, such a large population means lots of data from which to identify patterns. The ultimate goal is to correlate many data points in order to drill down to an understanding of smaller and smaller groups.
Hyper personalized
Hardoon spoke frequently of “hyper-personalization” and the ultimate goal of identify a data set of one – the individual, be it a person or a business.
For now, he says data is able to narrow groups to a few thousand people who are more likely to be receptive to an offer, such as a loan. The ability to do so makes it economic to serve low-income people or small businesses lacking a formal credit history.
The group is still laying the foundations. Hardoon described his first two big projects in this vein. First was to get business unit heads to think about how they could leverage the group to provide a compelling financial service. Second was to enable that.
Changing the business
From the business case perspective, ADI sits on top of the entire Aboitiz group. Finance is the easiest place to begin but ADI’s remit is to inject data science into all aspects of the conglomerate.
For example, Hardoon describes how better data can lead to predictions about the manufacturing of cement, so that product quality can be affirmed in a few days rather than a month – leading to operational efficiencies, faster output, and less wastage or energy usage. “There’s an ESG component to AI,” he said.
More sophisticated applications will emerge that leverage the data picked up from factory floors or distribution nodes to create financial offerings, such as supply-chain financing or invoice factoring.
These must be based on customer consent: someone gives UnionDigital or another Aboitiz arm approval to use specific data for a specific purpose. That’s key to embedded finance, in which customers may need a financial service to complete some other activity.
“Data allows the possibilities of a conversation [with a customer] beyond a specific business request,” he said.
Sharing data
This leads to the second of his foundation projects, a platform he helped build called Parlay. It’s an internal data-sharing platform in which data does not change ownership but is made sharable within a cloud-enabled environment, akin to the data rooms that broker-dealers use to trade.
The platform cuts out tedious paperwork for data-sharing agreements because the technology protects privacy and data.
Hardoon says the next step will be to offer Parlay to selected external partners.
He says the early use cases in finance are already being deployed, which will give UnionDigital a head start as other banks attempt to roll out their own AI-enabled services.
True deployment
“Everyone is doing AI and data science, but most companies are stuck in the proof-of-concept stage instead of deploying it in a way that makes a difference,” he said.
He ascribes that from the academic, wonky nature of data science, which doesn’t always communicate well with businesspeople. Secondly, AI is a process that takes time to hone, and can feel alien to banks relying on their traditional methods.
A market like the Philippines, however, offers big rewards for organizations that use data to reach big new customer segments. The laws of big numbers can give teams the confidence in their algorithms to start narrowing offers to more targeted groups.
“This touches on explainability,” Hardoon said, noting that ADI has to defend an AI’s decisions to regulators, customers or shareholders. He cited his experience helping MAS come up with guidelines for ethical AI to demonstrate his commitment to good governance.
Governance
Also: financial services are regulated. UnionDigital has to follow the same consumer-protection and other laws. “Fiduciary responsibility is baked in,” he said.
UnionDigital has also announced ambitions involving blockchain-based finance, including facilitating crypto transactions, NFTs, and DeFi services.
Hardoon declined to discuss the nature of the business cases, but says blockchain is a tool that falls under his remit. “Blockchain is also just data, with auditability and immunability,” he said. In particular it is a way to verify identities, which can be more efficient than traditional bank means. “As a former regulator, I’d love to do all ID checks on blockchain,” he said.
That said, he says the industry could use a data regulator. Hardoon asserts his commitment to principles such as transparency, trust, and user consent. But there are no formal guidelines for non-banks in the Philippines, and non-banks don’t have the same culture of compliance. Banks are regulated, but there are not regulations specific to fiduciary roles in terms of platforms or data sharing.
Hardoon says the tech itself provides a degree of trust, be it blockchain or a cloud-based service like Parlay. “Data sharing was the bane of my career,” he said. “Now trust is built in.”
