“For most organisations in Luxembourg, the cloud is now the preferred IT platform,” explains Ludovic Gilles, Chief Commercial Officer at DEEP. The question is no longer whether to use a shared platform, as was still the case a few years ago. It is more about which model to choose and the value the business can derive from it. “Today, companies do not necessarily wish to invest in their own infrastructure. With the cloud – whether public, private or sovereign – they have a scalable technological foundation, the operational management of which can be provided as a service,” continues Ludovic Gilles. “Above all, these platforms enable them to access the latest technologies and harness them to further their ambitions for transformation and innovation.”
Expectations are focused on business benefits
Through these models, the company can focus on issues other than the actual management of the IT infrastructure. “The organisations we support are primarily interested in the opportunities for transformation offered by the solutions we implement,” continues Ludovic Gilles. “Their full attention is focused on the potential gains in performance, productivity and competitiveness made possible by technological advances, particularly AI.”
For digital services companies such as DEEP, this therefore involves supporting businesses in their decisions, from consultancy through to the design and management of their technological environment. “The choice of infrastructure, therefore, must be made with the company’s ambitions and the constraints of its business in mind,” adds Ludovic Gilles. “When it comes to the cloud, we need to distinguish between the existing models without pitting them against one another.”
When it comes to processing highly sensitive patient data, a hospital may opt for an environment offering greater control, such as a sovereign cloud or a private cloud. Other functions, involving less sensitive data, may benefit from solutions based on the public cloud. Today, the sovereign cloud – such as the one DEEP has implemented with OVHcloud – primarily meets sovereign requirements, but also addresses the challenges of data control, security and technological independence faced by certain organisations.
Helping people make the right choices
Decisions must be made taking into account a range of factors, such as the location of the data, the levels of safeguards relating to the applicable jurisdiction, operational control or the reversibility of solutions, and, of course, the costs . Experience has shown us that, depending on the type of processing involved, the public cloud is not necessarily the cheapest option .
To weigh up the options, a company must therefore start with a robust risk analysis. With this in mind, DEEP aims to guide organisations, enabling them to maintain control over their environment whilst gaining access to new capabilities for transformation and innovation.
Deploying AI in an informed manner
With AI, these issues take on a new dimension. Today, access to the most powerful general-purpose models often involves the public cloud. However, depending on the business’s operational needs, it may be preferable to deploy more specialised open-source models on dedicated environments. “Once again, these choices will depend on the level of sensitivity of the data and the processing involved – in other words, how artificial intelligence is being used,” continues Ludovic Gilles.
For DEEP, the aim is to support businesses in a controlled and gradual adoption of the technology. “Initially, businesses approached AI through individual use cases, allowing staff to use assistants such as Copilot,” continues the Chief Commercial Officer “Then, we saw organisations exploring new use cases through pilot projects. The real challenge now is to scale up promising use cases whilst confirming their return on investment at scale.”
AI put to the test in the real economy
In this instance, the roll-out of AI does not always lead to increased profitability. “AI, today, must prove its value in the real economy. When certain use cases are scaled up, the resource consumption involved, and the associated costs, can dampen the enthusiasm generated by a pilot project,” explains Ludovic Gilles. The success of a pilot project alone therefore does not guarantee a return on investment.
Transforming the business, rather than simply optimising tasks
One of the challenges is to be able to estimate costs in advance, by identifying the most suitable model for solving a given problem or improving a process. “Today, companies are still focused on improving productivity. AI facilitates document research, document drafting, staff support and software development,” continues Ludovic Gilles. “The potential of the technology, however, should encourage us to look beyond the improvement of isolated tasks. In many companies, we have not yet fully reached the stage of business transformation. Yet that is where its true potential lies.”
This involves reviewing processes and the way data flows. With this in mind, DEEP is evolving the support it provides. “We are strengthening our advisory approach by focusing more closely on our clients’ business needs, as we have recently done on AML/KYC issues in the life insurance sector ", concludes Ludovic Gilles.
