Tech Talk: Under the Hood at GLBNXT – Sovereign AI Stack, On-Prem Deployments & AI-Driven Development
For many IT architects, platform engineers, and technical consultants, the rollout of enterprise AI remains a balancing act between speed, control, and infrastructure. How do you set up an AI stack that’s not only scalable and high-performing but also runs where your data needs to be—whether that’s on a Sovereign Cloud, an on-premises bare-metal cluster, or with a hyperscaler? During this Tech Talk, Jan Saan (co-founder of GLBNXT) takes a deep dive under the hood of the GLBNXT platform. No high-level marketing slides—just a pragmatic analysis of the underlying technology, the architectural choices for sovereign AI infrastructure, and the operational realities of running and maintaining such a platform. In addition, Jan will demonstrate in a comprehensive live demo how the entire software development cycle within the platform is driven by AI, and how to manage Day-2 operations (monitoring, model updates, and guardrails) across hybrid and air-gapped environments.
What will be covered during this Tech Talk?
- The GLBNXT Tech Stack & Architecture:A deep dive into the platform’s layers: from model serving (such as vLLM/TGI) and LLM orchestration to RAG pipelines, vector databases, and retrieval mechanisms.
- Flexible Deployment Models (Sovereign, On-Prem, and Hyperscaler):How the platform is built using Kubernetes to seamlessly deploy to a sovereign European cloud, proprietary on-premise hardware (with local GPU orchestration), or AWS/Azure/GCP without vendor lock-in.
- Live Demo 1: Full-Cycle AI-Driven Development:See live how software and AI assistants are developed end-to-end using AI agents, from code generation to integration and deployment on the platform.
- Live Demo 2: Platform Maintenance & Day-2 Operations:How do you maintain, monitor, and audit an AI platform in production? Focus on observability, latency, model evaluation, data isolation, and enforcing guardrails.
- Specific Client Use Cases:Real-world examples of how clients use GLBNXT for complex data workflows, including the architectural choices made in these implementations.
About the Speaker
Jan Saanis a co-founder of GLBNXT. Prior to that, he served as CTO at CM.com for over 20 years, where he built a globally scaled R&D and infrastructure organization from the ground up. With his deep expertise in distributed systems, platform engineering, and cloud infrastructure, he focuses at GLBNXT on building a robust, 100% sovereign AI stack for enterprise environments.
