Added business value without compromising on digital sovereignty

The rapid development of AI platforms and applications presents businesses with strategic and technological challenges. On the one hand, AI is seen as a key technology for entrepreneurial innovation and efficiency. On the other, there is the threat of new dependencies on proprietary technologies. It also comes with unforeseeable follow-up costs and entails a possible premature commitment to solutions and standards that could soon be outdated. This makes quite a few businesses hesitate to invest in AI.

A Red Hat survey of more than 100 Swiss IT decisionmakers and AI engineers has revealed that the largest group (31 percent) is still in the exploration phase. Twenty-seven percent stated that they were preparing for AI, while 23 percent want to raise awareness of the importance of AI internally. Only 16 percent reported revenues from their AI investment, and only three percent indicated that they were already generating added value for their customers from it. Digital sovereignty in the use of AI was a priority for 77 percent of respondents. “Enterprise Open Source” – open software specifically for companies – has been identified as a key technology for supporting managers on their path to value creation.

In collaboration with ti&m, Red Hat offers a solution that closes the gap between dynamic AI innovation and the specific needs and possibilities of local businesses. Global trends – above all the growing need for digital sovereignty as a result of geopolitical upheaval – must also be taken into account. So too must Switzerland’s specific market and compliance requirements. Red Hat’s specialization in open source solutions for the hybrid cloud lays a solid foundation which also allows manageable, targeted offerings through the consistent modularization of AI solutions. ti&m’s implementation expertise enables businesses to select the solutions that make sense for them, adapt these as required, and then implement them with the necessary control and flexibility.

After a phase of almost playful experimentation, the dominance of a few proprietary and mostly foreign providers in the AI sector has led to concerns even among Swiss companies about dependency, a lack of transparency, and a possible loss of control. However, Red Hat, which was founded in 1993 and emerged from the open source movement, has been able to transfer its experience with the open source model for enterprise needs to the AI stack. In particular, it offers four advantages that help businesses strengthen their digital sovereignty and minimize risks:

1. The ability to host, operate, and adapt models in-house ensures that sensitive data, including trade secrets, does not have to leave the internal infrastructure or Switzerland in general. This is particularly relevant for highly regulated industries.

2. Compatibility with the chosen model, including open models that allow businesses to review their own training data, architectures, and algorithms. This is a basic prerequisite for adherence to ethical guidelines, risk assessments, and the fulfillment of compliance requirements.

3. Open technologies encourage collaborative development. Community projects create the tools for businesses to fine-tune models with their own data, enabling shorter innovation cycles, greater relevance, and accuracy.

4. Open architectures offer flexible scaling. They help to limit the often unpredictable operating costs of proprietary AI services and make their budgets plannable.

Red Hat AI 3, a comprehensive refinement of the open Enterprise AI platform, was released in October 2025. Combining the Red Hat AI product portfolio (Red Hat Enterprise Linux AI, Red Hat OpenShift AI and Red Hat Inference Server AI) into a unified environment enables AI applications to move from proof-of-concept to production and amortization faster. It is based on open standards and supports AI workloads across hybrid multi-vendor environments, from in-house servers to the public cloud and the edge. Red Hat AI 3 supports any model on any AI accelerator.

Red Hat Enterprise Linux AI, the basic component, is an optimized, bootable distribution of the world’s leading Enterprise Linux platform. It bundles all the necessary components for the development, testing, and operation of LLMs. These include the open-source licensed “Granite Large Language Models” from IBM Research, which were developed for enterprise use cases and offer a trusted basis. The integrated open-source project “InstructLab” allows basic models to be improved easily and continuously. Domain and technical experts can also use it to refine and adapt the models with their specific knowledge.

Red Hat OpenShift AI, the centerpiece, is designed to operationalize the lifecycle of MLOps (Machine Learning Operations) on a consistent, business-ready basis. AI workloads, which typically place high demands on computing power and data availability, can be shifted seamlessly. OpenShift AI ensures that the development environment for data scientists (training) and the production environment (Red Hat AI Inference Server) are identical. This significantly accelerates the transition from development to production. Red Hat AI 3 also integrates a broad ecosystem of partner tools and open source technologies to cover the MLOps process from data collection to model monitoring.

ti&m is bringing its expertise in Red Hat OpenShift to this partnership. This will allow us to develop an AI strategy with customers that meets their actual needs. ti&m’s role today begins with strategic consulting and support: assistance in defining operationally relevant applications and an individual roadmap for the introduction of apps. The architecture design focuses on the joint concept design and implementation of the infrastructure based on Red Hat OpenShift, which meets the high data protection and performance requirements in Switzerland. Operational applications, specifically open source LLMs and the associated fine-tuning tools, could be integrated in the future. The combination of Red Hat’s leading open source technology and ti&m’s implementation expertise gives companies the control and flexibility they need to make the drive for AI innovation safer, faster, and sovereign at all times. This offers added value for current and future business without having to compromise on digital sovereignty. An expansion with Red Hat AI 3 could be an interesting platform alternative for ti&m’s customers.