Machine Learning garners impetus in Africa

Machine Learning garners impetus in Africa

To make decisions more quickly and accurately, enterprises in Africa are increasingly turning to Machine Learning, arguably today’s most practical application of Artificial Intelligence (AI). How should enterprises ensure success and ROI from Machine Learning deployments in their IT environments?

Machine Learning is a type of AI that allows software applications to become more accurate at predicting outcomes without being explicitly programmed to do so. Machine Learning algorithms use historical data as input to predict new output values. In addition, Machine Learning systems apply algorithms to data to glean insights into that data without explicit programming: It’s about using data to answer questions. As such, enterprises across the African continent are applying Machine Learning to a wide array of issues, from customer purchasing patterns to predictive maintenance.

According to research and consulting firm International Data Corporation (IDC), spending on Artificial Intelligence (AI) systems in the Middle East and Africa (MEA) is expected to maintain its strong growth trajectory as businesses continue to invest in projects that utilise the capabilities of AI software and platforms.

An IDC survey that interviewed over 2,000 IT leaders found that the adoption of Machine Learning increased customer and employee experience by 25% and also led to accelerated rates of innovation within the organisation.

Analyst firm Gartner has estimated that by 2022, 70% of white-collar workers will be interacting with chatbots daily.

Automotive vehicles are also able to collect complex data from their surroundings and interpret it to make precise and accurate decisions on their own, using Machine Learning. IDC forecasts that the number of vehicles capable of level one autonomy (driver assistance) will increase from 31.4 million units in 2019 to 54.2 million units in 2024.

Fady Richmany, Senior Director and General Manager – UAE, Dell Technologies, said with the developments being made in the field of Machine Learning today, the practical uses in enterprises are endless. Richmany said Machine Learning systems can be used to help anticipate trends and identify problems, thereby playing an important role in supporting decision-making processes. “Enterprises can also use Machine Learning for customer retention, since Machine Learning systems can study customer behaviour and identify potential steps for customer retention,” he said. “Additionally, they can make use of Machine Learning to help with market research and customer segmentation. This allows them to deliver the right products and services at the right time, while also gaining valuable insights into the purchasing patterns of specific groups of customers to better target their needs.”

He said furthermore, enterprises can also increase their operational efficiency by deploying Machine Learning to handle day-to-day routine business tasks, thereby speeding up operations, freeing up their employees for more innovation and creating new business opportunities.

Machine Learning secret sauce

With vendors often claiming to have some Machine Learning secret sauce in their wares that will revolutionise an enterprise’s business, CIOs are being urged to be careful when selecting the right Machine Learning systems and tools.

Alan Jacobson, Chief Data and Analytics Officer, Alteryx, said the three top considerations to selecting any Machine Learning system should always include usability, breadth of scope and an outcomes-based view.

Jacobson said much of the emphasis for Machine Learning has been on the technology, not the people and that’s where failed projects are rooted. “As technologies continue to converge, so do the consumers or producers of those capabilities. The current skills gap continues to be any issue for enterprises. There is a distinct lack of data scientists across the globe,” he noted.

Second, according to Jacobson, enterprises have to ensure that the chosen solution can help clean and manipulate data from whichever sources are necessary and operate across the breadth of the tech stack in place. “The end goal should be replacing any disconnected tools with hyper-specific functions in favour of one broad-use tool,” he said. “Finally, as with any technology purchase, an outcomes-based approach is also essential in keeping things on track. Is your organisation able to benchmark these outcomes in advance from existing use cases, for example? A focus on the long-term business impact and a direct impact on productivity are two key metrics to assess here.”

Rick Rider, VP, Applied Innovation, Infor, said endless compute resources is the reason Machine Learning is seeing wider industry acceptance now. In addition, Rider said it’s actually more about giving users the ability to use Machine Learning without building out or connecting technologies for years. “Now we have platforms that allow quick experimentation and implementation so efficient ROI is quite real with the right vendor,” he said. “Some industries tend to be more progressive, such as industrial manufacturing, distribution and more. However, even within certain industries it’s more about the companies who aggressively embrace new opportunities and technologies that succeed. Those are companies that tend to continuously find new ways to pivot and expand their business, regardless of industry.”

Getting business buy-in

Stephen Gill, Academic Head, School of Mathematical and Computer Sciences, Heriot-Watt University Dubai, said to remain relevant and competitive, a CIO must adopt two positions within their organisation: guardians of infrastructure and digital catalysts of business value. Gill said as Machine Learning and AI continue to transform businesses across a myriad of sectors, organisations are gradually starting to see their huge potential. “As with any initiative, stakeholder support is key for its eventual success and that is why so many CIOs focus on creating solid, evidence-based business cases for the technology investments that they want the management to approve,” he said. “While quantitative-empirical communications could be influential for other IT colleagues, CIOs also need to address non-IT stakeholders (especially members of senior management). They can do so by telling persuasive stories that illustrate the impact that the investment in emerging technologies such as Machine Learning will have on multiplying business value, especially on profits and revenue.”

He explained that being a good storyteller and a salesperson might not come naturally to a CIO who has risen through the IT and engineering ranks hence, it is important for them to develop such communication skills from their counterparts in sales and marketing, in order to gain the management buy-in and support needed to deploy their digital initiatives.

Dell Technologies’ Richmany said companies today know they need to increase their investment in new technology, however, they are hesitant of change. He said executives may not be fully aware of the benefits Machine Learning technology can provide to the business.

“Therefore, before making a case, CIOs and IT decision-makers must analyse the organisation’s vision and goals, and look for the IT solutions that will support in achieving these goals. They must make the case for Digital Transformation through a mission driven and business value perspective, which will allow key decision-makers to develop a better understanding of what the business is investing in,” he said. “This could include a focus on revenue generation, profitability, as well as employee efficiency and productivity. CIOs must also take security into account at every step of the way and have a solid security plan in their proposal.”

Challenges enterprises face

Priyanshu Vatsha, Intelligent Automation and Pre-sales Consultant, Proven Consult, said technical challenges associated with Machine Learning systems are majorly related to data. Vatsha noted that data unavailability, noisy, redundant or inadequate data makes it difficult to achieve satisfying results. “Problems also arise if input data is biased or encrypted. Ongoing validation is an additional challenge for the implementation of Machine Learning models in practice,” he said.

Vatsha said coming to non-technical challenges, building user trust is a big one. “Users need to rely on them when facing the challenge of making important decisions. Legal requirements also often pose a significant challenge for a Machine Learning project. This relates to data privacy protection as well as decisions on who is going to be accountable for false decisions based on Machine Learning models.”

Dell Technologies’ Richmany said Machine Learning is still in its very early stages and even at such an early development stage, the region is seeing it revolutionising a range of industries, with research and development advances being made in Machine Learning every day. “For enterprises to ensure the success and ROI of Machine Learning deployments, it is important for them to align them to defined clear goals and use cases, and associate these to business priorities. “Identifying and understanding whether the problems they are trying to solve could be tackled better and more accurately by Machine Learning rather than conventional software is key. Additionally, having experts run an elaborate experimentation phase of the potential projects which includes everything from gathering and assessing data, to basic modelling, cost and risk assessment can help predict whether the project will be successful or not,” he added. “This requires nurturing an organisation culture that values innovation. In the near future, we can expect quantum computing to significantly increase the capabilities of Machine Learning. It will give Machine Learning the capability to create systems that execute multi-state operations simultaneously. Quantum Machine Learning will have the ability to tackle complex issues in a split second.”

According to Jacobson, many organisations face challenges in moving Machine Learning models into production environments. On average, between 60% and 80% of models created with the intent to deploy are never deployed. “Plus, it typically takes six to eight months to deploy a model using legacy technologies, which leads to many projects becoming obsolete before they can go live,” he noted. “Machine Learning models are not a ‘one and done’ exercise. Model management, which can involve monitoring, revisiting and retraining, is a fundamental part of their life cycle. How you put your model into production will determine how easy it is to manage. Yet, even getting into production could be challenging.”

In fact, said Jacobson, this step was the highest hurdle cited in Davenport’s Machine Learning research, with 47% of executives saying that it has been difficult to integrate Machine Learning projects with existing processes and systems.

“Enterprises that struggle to integrate Machine Learning applications with existing production applications waste time and money on data science projects that are never put into production,” he said. ”Machine Learning operations (MLOps) is the critical process that makes this possible by treating ML and other types of models as reusable software artifacts. Models can then be deployed and continuously monitored and retrained via a repeatable process. As such, it helps businesses discover valuable information and insights from their data more quickly.”

Heriot-Watt University’s Gill said one of the most common Machine Learning challenges that enterprises face is the availability of data. “Proper access to raw data is very important as training Machine Learning algorithms requires huge amounts of data,” he said. “Insufficient data makes it harder to train the models properly which makes the implementation of Machine Learning project even harder.”

Gill said in addition to proper collection of data, enterprises also need to model and process the data to fit the algorithms that they will be using. “Another major challenge faced with Machine Learning implementations is data security. Distinguishing between sensitive and insensitive data is crucial for implementing Machine Learning correctly and efficiently,” he said. “Sensitive data should be encrypted and stored in other servers or at a location where the data is fully secured. Only trusted team members should be allowed access to confidential data.

Click below to share this article

Browse our latest issue

Intelligent CIO Africa

View Magazine Archive