ABA Technology showcases Fusion AI In-a-Box to Jensen Huang at GTC Berlin 2026
On the occasion of GTC Berlin 2026, ABA Technology presented its Fusion AI In-a-Box platform to Jensen Huang, CEO and founder of NVIDIA.
This presentation highlights a strategic ambition: to develop artificial intelligence infrastructures capable of operating locally, while guaranteeing data control, the security of sensitive information, and the technological autonomy of organizations.
An initiative that illustrates the contribution of Moroccan tech players to the global challenges of sovereign AI.
Vision & Positioning
Making sovereign artificial intelligence a lever for technological autonomy
At GTC Berlin 2026, ABA Technology, a Moroccan technology group led by Mohamed Benouda, presented Fusion AI In-a-Box directly to Jensen Huang, CEO and founder of NVIDIA.
According to an article published by Atalayar on October 24, 2026, this platform is among the five most promising initiatives identified by NVIDIA in the development of open models for European and sovereign public sectors.
This recognition comes at a time when organizations are seeking greater control over their digital infrastructure, data, and artificial intelligence processing capabilities.
The Fusion AI In-a-Box value proposition is based on a foundational approach: enabling businesses and institutions to develop and run their artificial intelligence applications within their own infrastructures.
From this perspective, sovereign AI represents more than just a technological evolution. It also constitutes a matter of governance, strategic independence, and control over digital resources.
For ABA Technology, this focus aligns with a vision where innovation must empower organizations to take full advantage of artificial intelligence while retaining control over their digital assets.
Concrete applications & societal impact
Infrastructure designed to support the entire AI lifecycle
Fusion AI In-a-Box offers an architecture covering various stages of the development and operation of artificial intelligence solutions.
The platform is built around two complementary configurations:
1. Fusion AI Factory In-a-Box – Development
This configuration is designed for the design and development phase of AI solutions. It notably enables:
Data preparation and management.
The adaptation and creation of artificial intelligence models.
The integration of AI solutions into business processes.
2. Fusion AI Inference In-a-Box – Production
This configuration focuses on running models in a production environment, with two main objectives:
Optimizing processing performance.
Reducing the operating costs of artificial intelligence systems.
Beyond these two configurations, the article highlights several components of the platform:
Advanced data management.
Support for multiple artificial intelligence models.
The creation of intelligent agents combining specialized software with NVIDIA computing infrastructure.
Outlook for organizations
This approach is of particular interest to public institutions and companies that handle confidential information or wish to integrate AI into their operations.
Potential applications notably involve process automation, data analysis, and the development of intelligent agents tailored to organizational needs.
The article presents the platform's capabilities without documenting specific operational deployments or quantified customer results.
The challenge, therefore, is to make AI infrastructure more accessible, manageable, and economically viable as organizations expand their use cases.
Sovereignty & responsibility
Mastering data to build responsible artificial intelligence
Data sovereignty is one of the central issues of Fusion AI In-a-Box.
Unlike architectures that rely exclusively on external cloud infrastructures, the platform prioritizes the local execution of artificial intelligence applications.
This approach allows organizations to retain their data within their own IT environments, without having to systematically transmit it to external vendors.
It addresses several strategic concerns:
Data protection: limiting the exposure of sensitive information to external infrastructures.
Technological autonomy: reducing certain forms of dependency on digital service providers.
Governance: strengthening organizational control over data and AI systems.
Operational efficiency: optimizing the resources required to run models.
For public institutions and enterprises, these dimensions are particularly important when artificial intelligence systems are involved in processing confidential information.
This direction also raises an essential responsibility: technological sovereignty must be accompanied by robust mechanisms for security, data governance, transparency, and human oversight.
From this perspective, ABA Technology advocates a vision in which innovation in artificial intelligence must go hand in hand with infrastructure control and the creation of sustainable value.
The development of sovereign solutions thus represents a strategic opportunity to strengthen local technological capabilities and contribute to a more autonomous digital ecosystem.
