BY Saarah Survé 2 MINUTE READ

Frans Cronje, Co-founder and Managing Director of DataProphet, left his job to develop a technical skill in artificial intelligence (AI). He had his sights set on monetising his newfound passion by either starting his own business or joining a team that were AI-focused. Thus, DataProphet was born. Along with the help of a friend, Frans noticed a gap in the market for a company that would consult and build machine-learning solutions. Later, they would gain enough knowledge to build and offer their very own product.

How has DataProphet used AI to propel the industry forward?

As factories have become increasingly mechanised, the amount of data that they produce has increased significantly, but the approach to analyse it has not changed. It’s still usually done in a very human, static manner (i.e. someone will pull data from a machine and spend days analysing it). It’s not uncommon for there to be much more data than someone could feasibly analyse and this is where AI comes in. It excels at ingesting and then interpreting the data, and providing a small set of the most relevant information to the operator. We developed OMNI with the same objective in mind, except it performs that analysis continuously, for every machine in the plant, knowing what is happening everywhere, and then providing that feedback to the operator before a defect occurs.

What makes DataProphet innovative?

By building a team that is interested in the development of the field of AI, and enabling them to bring that interest into their work, we continuously innovate.  One of the most interesting, difficult decisions we made was to move to deep learning as a sub-field in AI. This decision was in contrast to the typical 80/20 rule that is common in business.  Within AI, there is an older class of theory – consisting of a set of algorithms – that provide results that are often nearly as good as those produced by deep learning. Plus, they are often easier to implement. Early on in DataProphet’s history, we decided to focus on deep learning with the expectation that the newer field would enable us to innovate in the long term, and we have seen that to be true.

Read more in the March/April issue of Fast Company South Africa.