Around 4.3 million South African adults, or 11% of the adult population, are considered ‘financially excluded', a figure that rises to 66% in sub-Saharan Africa.

The World Bank cites documentation requirements as one barrier to accessing financial services on the continent. But, with solutions like robotic process automation (RPA) and facial recognition technology, this no longer has to be an impediment to financial inclusion.

That's according to Mike Fairon, sales and marketing manager at Yellow Dog Software, who says with the continent's high mobile penetration rate, there's no reason why financial institutions and the public sector can't serve remote and under-served areas.

"Using RPA and facial recognition technology, opening a bank account is as easy as taking a selfie on a mobile phone, with the entire onboarding process completed in under four minutes," says Fairon.

As simple as it sounds, Fairon notes there's a lot going on behind the selfie scenes to make this possible. While many bots and algorithms work around the clock to verify and validate new customers' identity, others run fraud checks and create paperless, secure profiles of every customer, quickly and accurately. That data is made available to other departments within an organisation to fast-track access to other financial services, like loans and insurance for small rural businesses.

Mobile opportunities

A few years ago, it was not feasible for someone living in a rural area to spend time and money travelling to a bank branch to open an account, only to be told they did not have the correct documents.

This prompted Africans to come up with their own savings mechanisms and stoked the growth of the informal savings economy, characterised by stokvels, credit unions and savings clubs. While effective, informal saving schemes do little to financially empower low-income earners or stimulate the economy. But Fairon says RPA could soon change that.

Today, rural residents only need a camera phone and an Internet connection to access financial services. Although connectivity is still a challenge in Africa, mobile phone penetration is among the highest in the world, at around 82%. This is expected to reach 100% by 2021 and smartphone penetration is expected to reach 80% by 2022. Sub-Saharan Africa also has the highest number of mobile money accounts. In 2017, 21% of adults in sub-Saharan Africa used their mobiles to transact, suggesting that mobile offers massive opportunities for the unbanked, and for service providers.

Lower costs, no fraud

Fairon explains that, when applying for a bank account, possibly at a mobile connectivity hotspot, new customers simply fill in their details, take videos/photos and upload their documentation, which is subject to rigorous fraud and validation checks through RPA. They're then required to perform a "liveness" assessment, where they're asked to do things like blink and open and close their mouths while looking into the phone's camera. Further security checks are applied to their live photo, which are correlated with their official ID document that the organisation already has on file or accesses through another data store. This makes it difficult, if not impossible, to dupe the system.

"The entire account-opening process can be automated, which drastically reduces a bank's acquisition costs and human resources, and leaves an audit trail for compliance and reporting purposes. Banks can use those cost savings to provide affordable financial solutions to the unbanked and reallocate their human resources to help those who are not digitally literate along the application journey," says Fairon.

Recognising diversity

The sophistication of facial recognition technology continues to evolve, says Fairon, to the point where it's now possible to verify someone's identity by scanning only their eyes and minimal surrounding facial features.

"This is important in a market as diverse as Africa, where we not only have people from many different races, but where we also need to respect the cultures and privacy of women who may not reveal their faces in public.

"As long as the organisation performing the identity verification can access an official document in their own data store or from a government data store, the selfie submitted through the facial recognition process can be validated in seconds, even if only a portion of the face is visible," says Fairon.

Endless applications

RPA and facial recognition technology has applications far beyond customer onboarding, says Fairon, citing voting in elections as another example.

"The government already has a data store of all our ID documents. With RPA and facial recognition, it's possible to log into an online voting system, scan our faces through our smartphone cameras and then RPA works in the background to verify our identity. Not only will this eliminate fraud, but the entire voting process can be done in minutes, rather than waiting hours in a queue at a voting station."

Government saves money not only on ballot papers and staff, but also because it can reduce the number of voting stations in urban areas, where smartphones and connectivity are more pervasive, and allocate more voting stations to rural areas.

Fairon says the applications of RPA and facial recognition are limited only by humans' imaginations. Technology has given us the power to drastically change lives, while saving costs, improving efficiencies and reducing fraud. We just need to get out of its way.