As Jaap Scholten, head, Group Hybrid ICT Strategy at Datacentrix and COO at eNetworks, a wholly owned Datacentrix company and ISP, explains, the cloud is often the unsung hero – quietly enabling data, the raw material from which intelligence is forged.

From cloud adoption to intelligent transformation

Too often, Scholten explains, cloud adoption is seen as a tick-box exercise, focused on moving applications from on-premises to the cloud.

“But the real opportunity here lies in transformation and at the heart of this is data. If cloud is the behind-the-scenes champion, then data is the golden thread weaving intelligence throughout the enterprise.

“At the end of the day, AI is only as effective as the data it consumes, and the cloud is essentially the ecosystem where this data converges. By unifying structured and unstructured data from across the enterprise, organisations unlock richer insights.”

Cloud powers AI at scale

For AI to deliver enterprise-grade value, it needs infrastructure that can handle unpredictable workloads, vast data volumes and complex integrations.

“Unfortunately traditional IT infrastructures can limit the speed, volume and flexibility required to make AI impactful,” Scholten notes.

“What is required is a cloud-native architecture able to provide the elasticity, connectivity and continuous data flows that allow AI to operate seamlessly across the business, detecting anomalies, trends and opportunities as they happen.”

While technology provides the foundation, the true measure of cloud-enabled AI lies not in the tools themselves but in the tangible business outcomes they enable, he adds.

Turning potential into performance

Yet, while the combination of cloud and AI unlocks enormous potential, success is not a given.

Organisations must navigate a set of practical and strategic challenges to realise meaningful value. These include reliable connectivity, safeguarding data, addressing cultural shifts and proving clear return on investment (ROI).

“These hurdles require deliberate planning and execution,” Scholten continues. “Only by tackling them head-on can businesses turn cloud-native AI into a sustainable competitive advantage.”

By aligning scalable cloud infrastructure, intelligent AI capabilities and a strong data-driven culture, businesses can unlock new levels of agility, innovation and resilience.

“Many companies might wonder where to start,” says Scholten. “The first step is a strong data strategy. Piloting AI use cases early and embedding governance throughout the cloud are also essential. Addressing challenges upfront ensures the transformation remains both secure and sustainable.

“The message is clear: organisations that embrace cloud-native AI today will define the competitive edge of tomorrow.”

Amazon Web Services launched its cloud services in 2006, followed shortly by Microsoft’s Azure offering in 2010.

Three years later, the word ‘hyperscaler’ entered our lexicon – meaning large cloud service providers that can provide offerings such as computing and storage at enterprise scale – and #CloudFirst became the buzzword amongst everyone, from developer to CISO.

Systems integrators were unsure how to measure the potential threat of cloud business, compared to traditional hardware and infrastructure sales. However, the groundswell of cloud adoption was not to be ignored, despite raising so many questions.

Business benefits and costs under the spotlight

One of the biggest shifts in executive think-tanks centred around ICT results versus business outcomes.

Spurred on by the COVID-19 pandemic, cost-saving efforts were being applied at all levels of the business, and ICT – known for its ever-increasing price/performance indexes – was put under the spotlight as an easy target for cost saving.

The question being asked was how a cloud-first strategy would align to business outcomes: was this a pure-play in technological evolution, which would only benefit the new economy – the Ubers and Airbnbs of this world? And would the cost of modernising mainstream businesses into a cloud-first era outweigh the benefit?

Early results posed questions

After witnessing the mass-migration of numerous large customer workloads, the first rounds of feedback were not as euphoric as the technologists had hoped for. More questions were raised, and yet another word entered our vocabulary: ‘bill-shock’.

Compliance officers also raised concerns over the sovereignty of company data. Patient records, student marks, financial information and intellectual property… where exactly was all of this data being hosted? And why do organisations have to pay to retrieve their own records?

As the dust settled, customers began moving some workloads back, in an attempt to regain control, both financially and in terms of compliance. A serious re-think of the cloud-first strategy was required.

Dollar-based billing resulted in IT budgets experiencing unprecedented cost increases, without realising any associated operational benefit.

The cost of extracting data, as well as the compliance issues around data sovereignty, rapidly led to a new approach. Given that almost all of an organisation’s records – customer, supplier, product and financial records, applying to entities both large and small – now lived as data somewhere, it became paramount to place data at the centre of such a strategy.

And so, the #DataFirst concept was born.

New solutions, better results: ‘Data First’

A healthy data-first approach results in a strategy that supports the fundamentals of where data is hosted, how it is transported, and how it is secured.

These underlying principles must be supported by a 360-degree approach, encompassing assessment, implementation, support, modernisation and continued gap analysis to assess the strategy’s execution progress. Ultimately, a data-first strategy is aligned to business outcomes and outperforms a pure ICT strategy.

Systems integrators started building smaller private/public clouds, hosted in sustainable data centres where power is guaranteed, with easily accessible sub-millisecond onramp paths and high levels of physical and cybersecurity, while addressing Rand-based billing and locally-based data sovereignty.

These clouds offer organisations Infrastructure as a Service (IaaS) as well as Platform as a Service (PaaS) options, which often mean a happy home for many applications that are not hyperscaler native.

With multiple availability zones, users of these services address their disaster recovery needs and can start to realise large-scale, long-term savings compared to their pure hyperscaler or on-premises deployments.

Systems integrators and cloud providers tend to concentrate top skills in these areas, thereby providing their customers with innovation, strategy, financial modelling and managed services all year long, while the customer can focus on their core business.

Multi-cloud adoption between different providers has proven to be both cost-effective and risk averse, now that multi-cloud management tools are readily available – even ‘as a Service’ – simplifying cost management, reporting engines, and optimisation efforts. Tools ensure that business outcomes are enhanced and realised.

As to the question “do we need another cloud”, the answer is therefore a resounding “yes!” – and there will be many more clouds to follow, almost moving into the boutique-genre of clouds designed for specific classes of workloads.

Data first, not cloud first

A critical first step when contemplating cloud services is that businesses must take a data-led approach, explains Scholten. “Data is an organisation’s gold, so to speak, and this should direct the decision-making process, not a particular technology. This makes the strategy business-led instead of being technology-led.

“Without a data-first approach, companies are unable to make relevant, real-time operational decisions and will end up being disappointed by the speed of their progress towards achieving their priorities.

“And by looking at where a company’s data should be, how it can be accessed, and how it can be secured, it becomes clear that it’s no longer justifiable to view technologies like cloud – or the network, or security – as individual siloes. They have all become inextricably intertwined and the data should dictate the technologies that should be weighed up.”

Physical infrastructure does not equal cloud power

It’s also important to shift away from the train of thought that there’s a like-for-like requirement when it comes to the transition from physical infrastructure to the cloud. “This is most definitely not the case”, explains Scholten, “in particular when you consider that physical, on-premises infrastructure is often not running at full capacity.”

Therefore, while cloud can be more cost effective, as there is no need for serious upfront investment, ongoing hardware, software and licensing updates, hosting fees or energy costs, it should still be noted that directly converting physical hardware into the cloud equivalent will be an expensive exercise.

Cloud’s consumption-based model ensures that there is no wasted capacity, as workloads can be scaled up as needed. This can be done quickly and easily, without the need for more hardware, software, labour and time that would be needed for scaling up on-premises solutions infrastructure.

“However, cloud isn’t and shouldn’t be thought of as the be-all-and-end-all of technology today, and there are several other factors that organisations should take into consideration when looking at a potential move to the cloud,” Scholten says.

“And while there’s no question that cloud adoption has been instrumental in initiating new business models, experiences and efficiencies in every industry, a large percentage of apps and data must still live in data centres and colocations due to issues such as latency, application dependency and regulatory compliance.”

The answer then, Scholten continues, lies in a hybrid IT approach: one that begins with a data-first strategy and provides an ‘as-a-service’ experience, delivering the agility and economics of the public cloud with the security and performance of on-premises IT.

Datacentrix model offers Rand-priced stability

Datacentrix’s cloud model is hosted in Teraco data centres and powered by HPE GreenLake. As a unique cloud offering and one of the first of its kind in Africa, a critical advantage of this approach is that it is available at a fixed Rand price.

“Aside from the clear multi-cloud experience benefits delivered, we are able to provide local currency stability to South African businesses, which is critical for those organisations going into the cloud that may be billed in dollars by hyperscalers,” adds Scholten. “By fixing the cost in Rands, we are able to remove the unpredictability of pricing.

“Our multi-tenanted approach has also opened up the infrastructure to companies of all sizes, meaning that there is no requirement for significant work demand in order to take the first step and enjoy the advantage of the many benefits offered by this cloud platform.”