From “Stargate” to Baltic Business
At the beginning of 2025, the United States announced Stargate, a joint venture worth approximately USD 500 billion, formed by AI giants such as OpenAI, Oracle, Nvidia, and others. The project promises a massive leap in AI infrastructure: supercomputer clusters, new data-center “cities,” and enormous computing power for next-generation AI applications.
At the same time, a critical question has emerged, also raised by Yale University researcher Madhavi Singh: does such a mega-alliance create risks for competition, pricing, and innovation when some of the largest competitors suddenly find themselves in the same boat?
This is not only an American story. Stargate highlights a broader trend in the AI era, infrastructure is becoming increasingly concentrated in the hands of a few major players. This directly affects Baltic companies as well, whose critical systems are increasingly dependent on the same four or five global suppliers.
Why “Stargate” Is a Signal, Not Just News from the US
According to Gartner, the three largest public cloud providers: Amazon, Microsoft, and Google control roughly two-thirds of the global IaaS (Infrastructure as a Service) market. In AI computing power, Nvidia’s GPU market share is estimated at around 80–90%. Stargate essentially unites parts of this infrastructure ecosystem into a single project. The Yale researcher points out that the risk is not only about today’s pricing, but about the erosion of independent “decision centers” companies that would normally compete with each other but are now coordinating their activities.
For a Baltic company, this means: if your entire digital infrastructure from ERP systems to AI services is tied to a single cloud provider or supplier, your negotiating power weakens over time.
The Single-Vendor Trap: What Happens on the Company Side
Juris Bergmanis, Head of Corporate Strategy and IT Projects at LTECH:
“Regardless of industry or business type, when I hear that a service can be provided by only one specific company, it raises concern for me as a consumer. We all value the ability to choose from several options based on different criteria, whether selecting a car or a TV as a private individual, or choosing telecom or utility providers as a business. The same applies to IT service providers.
In practice, however, we often see companies end up in situations where critical IT functions rely on a single supplier. Initially, a unified platform may seem convenient and logical, but in the long term the key issue becomes total cost of ownership (TCO) not only implementation costs, but also licensing, scaling, integrations, growth in users and data volumes, and the risk of cost increases over time.
Another important factor is technological dynamics. The IT landscape evolves rapidly, and companies must be able to adopt new tools and approaches when required by the market or regulation, not only when a specific vendor supports them. Otherwise, the vendor’s development pace begins to dictate the company’s ability to innovate.
Third, the deeper systems are built using vendor-specific components, the more complex and expensive migration to alternatives becomes. From long-term experience across different industries, we see a clear pattern: the longer a company remains in a single ecosystem, the higher the switching costs and risks.
This does not mean that a single vendor is always a bad choice. However, every organization must have an exit strategy. Without it, technological dependency becomes a high-impact risk that eventually materializes both in costs and in loss of flexibility.”
What This Means for Companies
- Large monolithic core systems tightly tied to one vendor or technology (e.g., specific databases, licenses, proprietary tools).
- High maintenance costs and technical debt that slow down new product development.
- Complex integrations, every new system must be adapted to one central platform.
- Limited ability to respond flexibly to regulatory or market changes, as every change becomes a major project.
Global studies show that outdated systems and technical debt cost companies hundreds of millions annually. In a recent survey, IT leaders estimated average losses of approximately USD 370 million per year due to legacy systems and failed modernization initiatives (ITPro data).
In the Baltics, the absolute numbers are smaller, but the proportions are similar: the deeper technical debt and vendor lock-in become embedded, the more expensive every future change becomes.
What This Means for Baltic Executives in 2026
Juris Bergmanis adds:
“Based on LTECH’s client experience and overall market observation, we see a clear trend. During modernization and digital transformation, companies aim for less manual work, higher system manageability, faster time-to-market, better data quality, fewer errors, and improved competitiveness.
Before starting cooperation, companies most often highlight several pain points:
- Inability to modernize IT fast enough without disrupting business operations,
- Lack of internal resources for technology upgrades,
- Concerns about rising costs and budget unpredictability,
- Slow delivery speed and limited flexibility from existing partners.
In the Baltic context, these challenges become increasingly critical over time. The longer technical debt and technological dependency remain unresolved, the more complex migration becomes, as data volumes grow, integrations multiply, and internal processes become more tightly coupled to existing systems.
This means that in 2026, executive leadership must focus not only on choosing the right solution for today, but on ensuring sufficient technological flexibility for the future. Only then can companies respond quickly to market changes, leverage innovation opportunities, and remain competitive long term.”
The Solution: Open-Source Foundations and Step-by-Step Modernization
Mārtiņš Kājiņš, Client and IT Project Manager at LTECH, sees an alternative:
“Long-term resilience means one thing, companies must remain independent in their technological foundations. This does not mean avoiding major cloud or AI providers, but designing architectures where replacing any single component does not require years and millions in investment.
Many open-source technologies already integrate well with commercial solutions. Instead of putting all eggs in one basket, companies can combine open-source components with commercial platforms, diversifying risk while reducing licensing costs.
Modernization is not a one-day or one-month effort. It is a structured, step-by-step process that gradually reduces technological risk and increases organizational freedom.
Step 1: A stable open-source database as the foundation.
In Baltic companies, the most common monolithic element is the database. Migrating to platforms such as PostgreSQL reduces licensing costs and allows flexible deployment, on AWS, Azure, Google Cloud, or on-premise.
Step 2: Unpacking the monolith, micro-steps instead of big-bang migration.
Modernization happens incrementally. Initially, one or two components with the highest business impact are selected and separated from the monolith through API layers. Business value appears early while risk remains low. For proof-of-concept projects, it is safer to start with components that have minimal business impact.
Step 3: Automation and test coverage.
This is where many organizations struggle. Manual testing makes modernization unsustainable. Automated testing ensures repeatability, safety, and long-term cost efficiency.
Step-by-step modernization is not only a technical project, it is a strategic approach to reducing dependency on global infrastructure providers and maintaining control and flexibility.”
Modernization as a Process, Not a “Big Bang”
LTECH supported one of the Baltic energy sector leaders providing natural gas wholesale and sales across Latvia, Estonia, Lithuania, and Finland in implementing a customer self-service portal. Previously, customers could apply for new services only in person or via paper forms by mail, a time-consuming process for both customers and internal teams. The portal introduced responsive design and secure authentication (bank ID, e-signature, Smart-ID). Customers can now create and manage service contracts digitally. Since launch in 2019, usage has grown from 20% to over 90%, significantly improving operational efficiency.
AI as a Layer on Top of a Well-Structured Architecture
AI should not be seen as a universal solution that improves any system automatically. Its effectiveness depends directly on data quality and system structure. When data is well organized, AI can identify patterns and deliver high-quality recommendations, leading to significant productivity gains. When data is fragmented or inconsistent, AI can generate misleading outputs.
According to Gartner, poor data quality costs companies approximately USD 15 million annually, making data governance a business risk management issue, not just an IT task.
What Baltic Companies Can Do in 2026
- Inventory critical systems and vendor dependencies.
- Assess technical debt and bottlenecks.
- Define 2–3 modernization priorities for 2026–2027.
- Select partners based on proven experience.
Conclusion
As Juris summarizes: “The AI era will not be about who launches a chatbot first. Winners will be those who maintain technological independence, even in a world where infrastructure is concentrated in the hands of a few giants. Open architecture and modernization are no longer nice-to-have; they are a form of insurance against future risks.”
Juris Bergmanis – Head of Corporate Strategy and IT Projects, LTECH
Mārtiņš Kājiņš– Client and IT Project Manager, LTECH
About LTECH
LTECH is a Latvian technology company founded in 2008, specializing in software development, system integration, and business process modernization. We help organizations improve efficiency by delivering secure, scalable, and user-friendly digital solutions that simplify processes and connect systems. Our team combines experienced engineers with next-generation IT professionals, fostering a modern, collaborative, and family-friendly culture.