The end of traditional software development and the need for hyperscalers?

For many years, software development followed a predictable pattern: teams focused on producing as many features as possible, as quickly as possible. This approach — often called a “feature factory” — ignored deeper questions such as Why do users need these features? How are we protecting data? and What legal rules apply when data moves across borders or organisations?

Today, these questions are more pressing than ever, especially in Europe, where regulations like the Data Act, AI Act, GDPR, eIDAS, DORA, and EUCS demand responsible data handling and secure digital services. Meanwhile, large-scale cloud providers—often referred to as “hyperscalers”—continue to offer enormous computing power and global reach. So, are hyperscalers still the way forward, or is there an alternative for organisations looking to balance compliance, security, and scalability?

Below, we explore how software development is changing, why hyperscalers have played such a big role, and how a node-based, mesh infrastructure is emerging as a viable option—either to complement or sometimes replace the traditional hyperscaler model.

1. The Problem with “Feature Factories”

1.1 Speed vs. Understanding In traditional software development, teams rush to add functionality, rarely stopping to consider whether these features truly meet user needs. Over time, organisations end up with bulky, fragmented codebases that are difficult to secure or adapt to new rules. For instance, if a sudden GDPR or AI Act requirement arises, developers are forced to bolt on compliance measures in an ad-hoc manner.

1.2 From Code to Context A better approach focuses on product logic — looking at why certain features are built and how data should flow securely and transparently. This product-logic mindset inherently satisfies legal obligations and user expectations from the start, rather than treating them as afterthoughts. Features become part of a larger narrative that ties together user experience, security, and data stewardship.

2. Why Hyperscalers Became So Popular

2.1 Scale and Convenience Hyperscalers like AWS, Azure, or Google Cloud solved a major challenge for companies: easy access to large-scale computing resources. Businesses no longer had to manage their own servers; they could instantly spin up virtual machines or databases, paying only for what they used. This was especially attractive for startups or fast-growing products that needed global reach and on-demand capacity.

2.2 The Challenge of Lock-In However, many organisations discovered over time that moving away from a hyperscaler — once deeply integrated — could be complex or costly. Certain hyperscaler services use proprietary APIs or data formats, limiting portability. Additionally, concerns about foreign jurisdiction or data privacy have grown, especially under regulations like the U.S. CLOUD Act, prompting a reevaluation of who truly controls the data.

3. Evolving Needs: Compliance, Security, and Local Control

3.1 European Data Regulations Europe has steadily increased its focus on data sovereignty. Laws like GDPR protect personal data across borders, while new frameworks — such as eIDAS for digital identities and the AI Act for ethical AI — raise the bar for transparency. These regulations highlight the fact that storing data in a single, centralised location might not always be the best option. In fact, certain industries must keep sensitive data within national or regional borders.

3.2 Zero Trust and Privacy by Design Security models are also changing. Zero trust architectures require every user, device, and service to prove its identity continuously, preventing the “insider threat” scenario where someone gains unrestricted access simply by being on the network. Privacy by design principles reinforce the need to safeguard data from the moment it’s created, not just after an incident occurs.

4. The Rise of Node-Based, Mesh Infrastructure

4.1 A New Way to Scale Alongside hyperscalers, a fresh approach has emerged: node-based, mesh IaaS (Infrastructure as a Service). This model uses multiple interconnected nodes, each potentially based in different regions or operated by different organisations, all woven together in a “mesh.” Unlike a single monolithic cloud, each node can specialise in local compliance requirements (e.g., quantum-safe encryption, national data residency) and still collaborate with other nodes when needed.

4.2 Data Space Architecture A mesh architecture also supports data spaces—decentralised networks where entities share data on equal terms. Information can flow one-to-one, one-to-many, or many-to-many, but always under strict rules and cryptographic protection. No single participant wields undue influence over the network. This federated design aligns well with European sovereignty goals, as it allows smaller organisations and public institutions to interact on par with larger companies.

5. Balancing Hyperscalers and Mesh Infrastructures

5.1 Complement Rather Than Replace Hyperscalers can still be ideal for global workloads requiring massive scale or advanced services like AI training clusters. Meanwhile, a node-based mesh is better suited to scenarios needing local compliance, specialised data handling, or guaranteed sovereignty. Most organisations won’t choose only one or the other. A hybrid strategy can pair hyperscalers for big compute tasks with local nodes for sensitive or regulated data.

5.2 Reduced Complexity, Better Compliance Organisations adopting mesh IaaS often find compliance easier because the architecture enforces data protection at every node. Workflows become more transparent: you always know where your data is and who can access it. This reduces the risk of hidden vulnerabilities or overlooked legal obligations.

6. TARI² and the Evolution of Product Logic

Traditional back-end development often demands coding complex workflows, compliance rules, and data pipelines from scratch. TARI²—Tasks, Activities, Rules, Information, Integrations — offers a smarter path. Instead of implementing the same logic repeatedly, TARI² uses structured metadata to define how data should flow, who can access it, and which rules must be enforced at each step.

This metadata-first approach drastically cuts down on development effort. Core capabilities—like orchestrating tasks or enforcing regulations—become reusable “blocks” that tenants can adapt to their specific needs. It also paves the way for AI agents to learn contextual data patterns. Rather than writing Word documents or building Excel reports manually, these agents can generate them automatically based on TARI² definitions.

Email and chat can also move into the same product context, ensuring all participants share the right information at the right time. Because TARI² defines who sees what (and under which rules), communication remains synchronised with real-time data. Everyone accesses the same structured context—no stale attachments or fragmented email threads.

Ultimately, the heavy lifting of traditional back-end coding begins to fade. Metadata orchestration, not raw code, drives most of the intelligence within a system. People remain essential for design and human interaction, shaping how services look and feel. But the foundation—compliance, data sharing, and process automation—relies on TARI²’s product logic. This shift frees developers to focus on innovation, helps organisations adapt faster to changing regulations, and ensures AI can operate transparently, always grounded in equal information positions and structured governance.

7. Real-World Advantages

7.1 Less Friction, More Trust Using zero trust security and TARI² in a mesh environment often reduces the hassle of multiple logins, scattered dashboards, and clumsy compliance add-ons. Users log in once, and the system verifies their identity in the background. The data they see is precisely what they’re allowed to see — no more, no less.

7.2 Empowering Local Innovation When smaller organisations have equal access to cutting-edge infrastructure, it broadens the range of what they can build. Instead of paying for large, one-size-fits-all services, they can tap into node-based solutions that match their needs. This fosters healthy competition and keeps data and innovation closer to local communities.

8. What Does the Future Look Like?

8.1 Hybrid Cloud Landscapes We’re likely to see both hyperscalers and mesh-based IaaS thrive, as different workloads call for different solutions. Many companies will pick “the best of both worlds,” migrating sensitive processes to secure, federated nodes while maintaining large-scale tasks on hyperscalers.

8.2 A New Development Mindset On the software side, developers are shifting from “just ship code” to “design with purpose, trust, and compliance in mind.” Product logic encourages them to build features that users actually need, while TARI² and zero trust ensure those features meet evolving legal demands.

8.3 Trust Through Transparency Ultimately, success depends on trust. Whether data resides in a hyperscaler data centre or a local node, users and regulators want proof that it’s handled responsibly. Federated mesh architectures and robust compliance frameworks can show precisely where data lives, who can access it, and why it’s needed—transforming trust into an everyday reality rather than a corporate buzzword.

The future..

Traditional software development — focused on rapid feature delivery without sustained thought for security or compliance—no longer meets the rising demands of an interconnected world. Hyperscalers have provided a powerful stepping stone to flexible cloud computing, but they’re not always the best fit for data sovereignty or regulatory complexity.

A node-based, mesh infrastructure, by contrast, offers a complementary path. It gives organisations the agility to meet local requirements, integrate compliance into workflows from the start, and foster true decentralisation with data spaces. Meanwhile, large-scale hyperscalers remain valuable for tasks that need massive global reach or specialised services.

In short, this isn’t an either-or debate. The future likely belongs to hybrid models, where teams balance the raw power of hyperscalers with the precise control of mesh nodes. By combining product logic, TARI²-style compliance, and zero trust security, organisations can build software that is both innovative and trustworthy, forging a new era of digital services that respects users, regulations, and the endless possibilities of the cloud.

Hans van Bommel

Founder of EU FED Cloud