OPINION: How to successfully apply AI in your business

Harkrishan Singh – Director (Application Development) at In2IT Technologies. Photo: Supplied

Harkrishan Singh – Director (Application Development) at In2IT Technologies. Photo: Supplied

Published Sep 9, 2020

Share

DURBAN - The term Artificial Intelligence (or AI) has existed in the realm of science fiction for decades, typically represented by intelligent humanoid robots.

As AI has become a reality over the years, however, it is important for businesses to realise that ‘artificial intelligence’ does not necessarily mean ‘artificial human’. While this type of robot is certainly becoming more feasible, they are simply one element of a multi-faceted technology that has almost endless potential, but also significant opportunity for failure.

Successfully applying AI to your business means firstly understanding the various use cases available. Only once you appreciate how it can be used to your benefit, can you select the most appropriate application, while ensuring the implementation has been effectively planned in order to minimise any potential pitfalls.

Use cases for maximum ROI

While AI has practically unlimited potential, its usefulness to business requires the return on investment to be worthwhile. Some applications are therefore more suited for AI than others, for example chatbots, which use Natural Language Processing (NLP) to interpret words and data to apply contextual and reasoning algorithms. This is useful for customer self-help as well as for other applications such as running surveys.

AI can also be applied to content distribution on social media, as it has the ability to enhance search capabilities, categorise products according to their attributes, detect inappropriate content, and more. This is helpful in assisting businesses to target the right customers with customised content, for example. Security, in particular cybersecurity threat detection and prevention, can be enhanced by AI. Some aspects include the ability to detect attacks in real time, analyse network traffic to detect anomalies, create sandboxes for malware, and improve endpoint detection and response.

Advanced analytics capabilities such as machine learning can create new AI capabilities such as preventative maintenance tools. This often utilises sensors and external data sources to manage the health of assets, alerting when potential issues crop up so that they can be dealt with before a failure occurs. In addition, predictive tools can be applied to areas like inventory management to help optimise the supply chain, and in quality inspection and assurance to automate inspections.

A solid foundation is essential

As with any new technology implementation, planning is crucial to the success of AI. Once an appropriate use case is selected, a number of factors need to be considered. First and foremost is data storage. AI applications generate vast data volumes, particularly when they include sensors and connected devices, and this volume is likely to grow exponentially. The ability to scale storage as the volume of data grows is therefore a critical consideration.

The networking infrastructure is another key component to successful AI, as connected technologies require high efficiency and availability in order to perform effectively. The network also needs to be able to scale as required to accommodate growth in data traffic.

Preparation is also key to success. Data sets that are to be used by AI algorithms need to be properly prepared and ensuring data integrity and data quality is essential. In order to generate meaningful insight, data must be accurate and coherent, otherwise analysis will inevitably produce irrelevant answers. This in turn means that data management and governance become critical, and data access controls must be in place to mitigate potential privacy and security issues.

Finally, as with any technology endeavour, training is key. Without the appropriate skills development internally, users will likely be resistant to the technology and adoption rates will be low. This then has a negative impact on ROI. However, empowering resources will skills and know how is easily achieved by partnering with a technology specialist.

A host of business benefits

When applied successfully through an appropriate use case, AI can have many benefits for business, from enhancing customer service through quick responses that overcome language barriers, to improving productivity and increasing efficiency through process automation.

AI can also improve the decision making process, create new business opportunities and increase operational efficiency, reducing human error and saving costs by optimising workforces and business processes. Many repetitive manual tasks can be delegated to AI, instead of requiring human intervention, and the back office function in particular can be dramatically improved with the assistance of AI.

However, arguably the most important benefit of AI is the ability to boost innovation, by analysing huge volumes of data rapidly, reducing the cost of research and development and creating new possibilities for experimentation. Ultimately, AI creates a new way of thinking about technology, business development and strategic execution, which, if effectively implemented, has knock on benefits for the entire business.

Harkrishan Singh – Director (Application Development) at In2IT Technologies

BUSINESS REPORT ONLINE

Related Topics: