Fraud detection is killer app for AI, analytics in Africa
Fraud detection is the primary driver behind analytics and AI adoption in Africa.
This emerged from discussions at SAS Analytics Experience 2019, the business analytics software vendor’s annual event, which was held in Milan this week.
Desan Naidoo, VP of Africa at SAS, said Africa faces similar challenges as its international counterparts when it comes to AI solution adoption – but the continent does differ when it comes to the adoption rate.
According to the International Data Corporation (IDC), growth in spend on AI systems in the MEA region is expected to reach $374m in 2020, up from $261.8m in 2018 and an anticipated $310m in 2019. The retail and finance sectors are expected to lead and, together, account for over 33% of the spend, followed by governments and the telecoms industry.
Naidoo is in charge of growing SAS Analytics’ software market share throughout Africa, and believes the maturity of South African clients is generally higher than clients on the rest of the continent. This is largely due to the fact that South Africa holds most of Africa’s data analytics skills.
Clients in the banking and telecoms sector in particular have had data analytics and advanced analytics skills developed in-house over many years. They are now transitioning to explore opportunities in open source, he adds. Programmes to develop those skills at universities have been there too.
“Fraud has been the number-one use case right now and we can also segment that into South Africa and the rest of Africa. In South Africa, we went through ten years of a government fraught with fraud and irregularities. The new dispensation is talking about actually fixing this.”
Naidoo thinks it’s time for our government to put its money where its mouth is as the technology that can detect fraud exists. “You need to embrace it…and make sure you’re implementing systems and processes to fix the problems,” he says.
SAS has partnered with a South African cyber forensics and data analytics firm called Facts Consulting, to develop its Procurement Integrity solution that tightly integrates procurement systems with analytics and looks for anomalies in that procurement.
The software scrutinises companies that win tenders by analysing various sources of corporate data and ownership records, looking for links between people within the company and those awarding the deal, and even uses image recognition to trawl through social media.
The technology has been integrated into SAS’ global solution portfolio and is expected to gain ground in Africa, specifically as businesses focus on fraud detection and prevention, rather than post-incident investigation.
“We’re seeing adoption within the banks right now, and not as much in the public sector – although with the financial intelligence unit, there are some discussions happening,” says Naidoo.
The second biggest use case for data analytics in Africa is customer service.
“Getting it right with the consumer is extremely important today,” he adds.