Globally, companies are indicating that they will invest even more in artificial intelligence (AI) in response to the COVID-19 pandemic and its acceleration of digital transformation initiatives. But while the potential of AI is clear, many local businesses still struggle to see how to effectively integrate this technology into their broader operational processes.
Part of this can be attributed to the view that the likes of AI and machine learning (ML) are just products that can be installed and forgotten about. But like any technology, these innovations can only truly become useful if they are embedded in a system where the people inside the organisation can use them effectively.
An example of this is the work Synthesis has done for Nedbank Insurance. Previously, employees at the financial services provider spent eight hours a day reading and categorising emails. But by introducing sophisticated AI and ML features as part of an overarching software development project run by Synthesis, a robot now does this 24×7 to make sure clients get a quick response. This has seen Nedbank Insurance clear its main client inbox almost eight times faster, getting rid of the backlog and speeding up client service.
Similarly, Synthesis used ML to speed up the Absa and Barclays brand separation in 2018. With a limited period to identify and replace the logo in key documentation, Web sites, contracts, database structures, marketing, and communications. It would not have been possible to be reliant on human-driven processes. Synthesis used an agile approach to deliver a software-based solution using ML and a micro-services cloud-based architecture to rapidly identify 40 000 artefacts that needed replacing.
Part of a whole
It comes down to applying AI and ML models to software development processes that look at business operations in its entirety. Being able to leverage the high-performance computing capabilities of cloud providers like Amazon Web Services (AWS) for example, AI and ML have the potential to positively disrupt any organisation.
But to do so requires several elements. Firstly, data scientists need to draw up the right model for a particular business objective. The data engineers then need to build that model and ensure data is processed correctly. Finally, it requires AI and ML to be applied to a modern, open-cloud framework that can scale securely according to the business requirements. AI and ML is not a bolt-on service but must be architected into a solution capable of delivering the advanced value needed.
While these technologies can also help in decision-making processes, the model used must be capable of learning from the human elements in the organisation. Human intervention at the early stages of the model development is crucial to ensure there is an effective feedback loop and an understanding of how the machine-driven insights are generated.
Ultimately, it is not just about the technology, but the entire business approach involved. AI and ML are not stand-alone concepts but must be built into an effective use case. Of course, being able to implement more sophisticated AI and ML opportunities depends on the level of maturity of the organisation to the technology.
But with every business wanting to improve efficiencies and maximise the potential of employees, AI and ML can help unlock value previously left untapped. With significant opportunities for these technologies across sectors in South Africa, companies are well primed to start taking advantage of what AI and ML can offer. However, they must remember to take an integrated approach to these technologies if they are to be truly effective.