PPS innovates with AI, ML to provide more personalised services for its members
PPS is reinventing its traditional legacy business through a digital transformation drive based on progressive leadership, strong technology partners as well as a change in organisational culture.
PPS's latest projects include harnessing artificial intelligence (AI) and machine learning (ML) to create a new platform that provides members with personalised services such as product recommendations.
"AI is not a singular concept, but rather a toolbox of different techniques that can be used to solve many types of problems," says PPS Chief Technology Officer (CTO) Avsharn Bachoo. "For our purpose, AI is the ability of computers to learn, gain experience and then make decisions just like humans, without being programmed, while ML is a subset of AI, which focuses on looking for patterns in data to predict the future."
Bachoo says most platforms around the world use techniques such as Bayes, clustering or inferences, but PPS has built its platform on Google Cloud, using AI and ML, called TensorFlow and Keras. The platform identifies patterns, uses information from the credit bureau as well as from PPS's own systems.
"Based on individual behavioural history patterns, our new platform has the ability to recommend genuinely personalised products to clients. It took us four months to build, and has already improved sales considerably within the first few months," says Bachoo. "We now use TensorFlow, which is Google's open source ML platform, and a game-changer in the industry. Earlier approaches such as clustering are often flawed because they are based on correlation, which often doesn't reflect causation. As with other recent projects, we have achieved this with the support of Google and Siatik, Google's premier Cloud partner in Africa."
"People ask us how PPS managed to do this so quickly," adds Brett St Clair, CEO at Siatik. "Let's take the analogy of a bakery. Imagine hiring a pastry chef to build you an oven. When it comes to ML, that's exactly the kind of mistake most organisations make. If you're opening a bakery, you want an experienced chef, and you obviously need an oven, but you won't ask your chef to build you an oven. Most organisations focus on the wrong thing: they focus on the algorithms, which is the equivalent of the oven. There are actually thousands of algorithms freely available. There is no need to reinvent the wheel. But many make the mistake of trying to build ovens from scratch. Those ovens exist already. Providers like Google let you use their entire kitchen, complete with appliances, and recipe books."
Bachoo notes most South African businesses have not been brave or bold enough to implement AI and ML projects. "But the reality is that your team doesn't need to understand the intricacies of computer science, maths, statistics or calculus. Don't fall for the lie that says you need expensive platforms, or data scientists with PhDs to do amazing things with ML," says Bachoo. "And there is no reason to hold back, given that ML gives you the ability to rapidly look for patterns in behaviour. It's not simply about pushing sales, it can strengthen your relationships with your customers."