Those who want to be less dependent on American software can of course switch to European alternatives. The same goes for contact center software: there is plenty of choice. But just replacing business applications is not enough. The IT stack of the contact center also consists of many supporting components. Insight into this complexity is necessary to be able to make a plan B. Ziptone helps you on your way.
Reducing this dependency on US software is relevant for two reasons. First, it makes the European economy less vulnerable to interventions such as tariffs or restrictions. Second, the current transatlantic agreement on data exchange between the EU and the US is under pressure. If that agreement expires – an idea the US has been toying with for some time – European organisations will no longer be GDPR-compliant. They will then have to enter into additional legal agreements – or they can limit the use of US cloud solutions.
However, the effect of the GDPR is not limited to your CRM system, a database or the customer contact system. On the contrary, the privacy legislation works through almost all layers of the cloud stack that companies - and also contact centers - use.
From one cloud to another: no piece of cake either
The cloud stack is a structure consisting of a large number of components, necessary to make the cloud environment work (see box below). The components are coordinated with each other and are specifically developed for a specific ecosystem – think of the cloud of hyperscalers such as Google, AWS and Microsoft. Their ecosystems contain all kinds of services and tools such as managed databases, provisioning tools, tools for orchestrating containers such as Kubernetes, solutions for object storage or automatic scaling. All of that makes cloud-based software work.
This toolbox is not only ecosystem-specific, but almost always also service provider-specific. IT service providers not only standardize their services, but also like to add – despite available best practices – self-made functionality with which they can serve their customers even better than the competition.
For example, Capgemini's Amazon cloud is not the same as Solvinity's Amazon cloud. It does mean that a choice for a specific cloud ecosystem (and sometimes a specific IT service provider) has a strong impact on all parts of the cloud stack. This phenomenon also occurs at application level: you cannot 'move' a CCaaS solution from AWS to Azure or vice versa, although there are SaaS providers whose solution is suitable for various hyperscalers.
There are known examples where contact centers from on-premise to a cloud solution in a short period of time have gone. But change So moving to another cloud is a different story.
Cloudstack, what is it? – Cloud (definition) is a collective term for the range of services and resources such as servers, storage, software and networks via the internet. The cloud stack – the collection of technology that makes the cloud work – is made up of all kinds of hardware and software components. This ranges from servers (hardware, with processors, memory, storage space and network connections – here too, American providers are dominant) to storage capacity for data, switches, routers and firewalls to send data over networks and solutions that allow multiple 'virtual servers' to run side by side on hardware.
Each cloud has different types of software on board, such as operating systems (think Linux or Windows Server) and virtualization solutions. Middleware is also needed that regulates the communication between the operating system and applications, such as web servers (Nginx, Apache) or databases (such as MySQL), and software for programming languages such as .NET, Java, Node.js, Python or PHP. Software is also needed that regulates things like security, authentication and identity management (of users and services).
Also, solutions like Containers are used to bundle all the components needed to run an application and Kubernetes to manage a series of Containers. And the last layer is the application itself, which is made available to the end user. In the contact center, you can think of the contact center software (the CCaaS-solution) that customer contact professionals work with.
For each layer in the previous list, IT providers can choose from different technologies to make that layer work. American solutions are also dominant in these technologies – and mutually coordinated. For example, Microsoft's IaaS environment (Azure) works well with both Docker's virtualization layer and Microsoft's SQL Server database solution.
Much of the cloud stack is relevant to GDPR
When you engage a cloud provider, you do not have control over a large part of the cloud stack. However, that does not release you from the obligation to maintain an overview of the technology that the cloud provider uses. This is necessary, among other things, to be able to properly assess risks and demonstrate compliance when using software.
For the GDPR, almost all layers and components from the cloud are relevant:
- Data centres: determine the physical location of data; data must remain within EU borders or in countries with equivalent levels of protection;
- Servers: this is where the actual processing and storage of personal data takes place, with requirements for encryption, access control and logging;
- Storage systems: subject to requirements such as encryption, data minimization and deletion capability;
- Virtualization applications: provide separation of software processes and are important for data security;
- Management solutions: such as Kubernetes for managing containers, impacting the way data is manipulated;
- Network components: such as firewalls and routers that secure and encrypt data streams;
- Databases and identity management systems: regulate access to personal data;
- Monitoring and backup tools: contribute (reactively) to the security and availability of data.
Processor or controller?
Often, a cloud service provider is also a data processor: personal data is processed that is stored in databases and file storage by the recipient of the services – the controller. If the service provider is given permission and/or instructed to use the data for other purposes, it is itself the controller. This could involve a CCaaS provider that trains its own AI tools with conversations that are held at the contact centers of software users. CCaaS providers may also work with subcontractors.
What applies to cloud ecosystems also applies to large providers of cloud-based contact center software (CCaaS solutions). They also have their own ecosystems, often in the form of a wide range of apps, plugins and APIs. Many of these are based on AWS or Azure standards and therefore not usable in an alternative cloud environment. There are, however, suppliers who have developed services for these cloud interoperability.
Challenges
The complexity and interconnectedness of cloud environment components in which cloud applications run pose multiple challenges.
First of all, as mentioned, there is no such thing as cloud portability. There is still little experience with (problem-free) moving of applications and data from one cloud to another. Providers and customers can take measures for this themselves. And there are various developments in this area, such as Gaia-X. In view of geopolitical developments, cloud portability is coming higher on the boardroom agenda from both the demand and supply side of the market. There are also projects in development that should ensure that applications are separated from (cloud) infrastructure.
Secondly, when changing the cloud environment, the entire operating environment must be re-arranged. This not only costs money, but also time. Time is a serious factor, for example, time is needed for the cloud provider to prepare (extra) hardware, migrate data and/or applications, set up the functionalities of cloud applications and integrate and link applications.
Thirdly, the ecosystems of European cloud providers are not yet as developed as those of the hyperscalers. Almost all the necessary components are available, but not always well integrated within the ecosystem. The gaps are mainly in highly specialized solutions.
In addition to these gaps, a move to another cloud platform can also lead to more complexity of your alternative cloud stack and a larger number of suppliers. Also, not all alternatives will be equally scalable and user-friendly when compared to the services of hyperscalers. On the other hand, non-US solutions offer more control over data, although you cannot rule out that US intelligence services will also enforce access to data here.
And finally, not only applications need to be moved, but also data. So-called metadata play a role in this and in many cases they are the property of the cloud provider. Furthermore, data structures differ per cloud environment. Retrieving data is also a risk factor. For large datasets, extra bandwidth is required and egress costs are almost always charged for this (for 'outgoing' data traffic).
Due to these challenges, in practice, moving a cloud stack will often take place in phases.
Creating – From CCaaS providers to system integrators, most of the parties Ziptone has spoken to in recent weeks are aware of the problem. There is also scepticism about the issue of dependency and the potential for acute vulnerability – neither the EU nor the US are ‘pulling the plug’ on the Data Privacy Framework', is the idea.
Read more – The reporting on developments in the US is relatively one-sided in Europe. Those who want more insight can read the updates from Kirsten Verdel about measures of the American government or read the NYT (free with a subscription to NRC). For those who want to know more about cloud migrations: cloud architect Bart Veldhuis of Rapid Circle (a Microsoft partner) has listed various IT considerations in this podcast.
Finally
Dependence on US hyperscalers entails real risks. There is every reason to reassess these risks now. Switching to a purely European IT landscape, either in whole, in part or in phases, requires thorough preparations and careful consideration. Essential questions are:
- Are our applications built with cloud portability in mind?
- Do we have insight into where exactly our data is located?
- What alternative software vendors and solutions are available to us?
- Have we formulated a good exit strategy?
There are sufficient European solutions available, for example as an open source solution, for the different layers and components of a cloud environment. In any case, now is a good time to investigate possible alternatives and map out the impact if alternatives have to be chosen (in whole or in part).
Additionally, companies can take proactive risk mitigation measures, such as implementing local data backups, setting up alternative application capacity and, where possible, storing data locally.
(Ziptone/editors)
Also read part 1 of this diptych: What is the alternative to big tech's cloud?
Featured, Knowledge base, Technology




