Juan Carlos Garcia Pelaez
Platform Engineer | Sovereign Agentic AI on Kubernetes
Barcelona, Spain
Actions
With 25 years of experience engineering core infrastructure for global organizations like the European Commission, UST or Atos, I have a strong connection to the Swiss tech ecosystem, having architected critical platforms for Roche (Basel) and Hexagon (Zurich). Today, through my company Kubekub, I tackle the industry's most complex challenge: securely operationalizing Agentic AI at scale.
I partner with engineering teams to bridge the gap between AI experimentation and production reality, with a specific focus on helping organizations technically comply with the new EU AI Act. I build hardened Cloud Native platforms that enforce fine-grained authorization, isolate autonomous workloads, and provide deep observability. I bring the discipline of a veteran Platform Architect to the bleeding edge of AI—I don’t just draw the diagrams; I build the platform.
Area of Expertise
Topics
EU AI Act: Securing Agentic AI with Platform Engineering
The EU AI Act introduces strict requirements for "High-Risk" AI systems, including deep traceability, fine-grained access control, observability, and process isolation. Failing to prove exactly what an autonomous agent did—and why—is a massive compliance liability, with fines reaching €35 million or 7% of global revenue.
While application code will always play a role in AI safety, Platform Engineering can shoulder the heaviest regulatory burdens by default. How do we build an Internal Developer Platform (IDP) that adds as much security and traceability as possible before the developer even writes their first prompt?
In this session, we translate the law into Cloud Native infrastructure. We will map the Act's requirements to a 100% open-source stack, demonstrating how to build a "Compliance-Ready" architecture. We will explore Kubernetes for agent sandboxing, standard fine-grained authorization (like AuthZEN) for data privacy, and the Model Context Protocol (MCP) for standardized tool execution. Crucially, we will demonstrate how to leverage AI Gateways to enforce strict guardrails and emit immutable execution traces, ensuring perfect auditability outside the agent's control.
The session culminates in a live demo: abstracting this regulatory complexity into a single Custom Resource Definition (CRD). You will see an AI agent deployed with strict, infrastructure-level guardrails—securely interacting with sensitive enterprise data systems—using just a simple kubectl apply. Attendees will leave with a vendor-agnostic blueprint to build an IDP that empowers developers while ensuring the platform is secure and compliant by design.
The EU AI Act introduces massive compliance risks and up to €35M in fines for autonomous AI systems. Instead of relying on application code for safety, learn how Platform Engineering can shoulder this regulatory burden by default. We will translate the law into a 100% open-source Cloud Native stack using Kubernetes, AuthZEN, MCP, and AI Gateways for immutable tracing. The session culminates in a live demo: deploying a secure, compliant-by-design AI agent via a single CRD with just a simple kubectl apply.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top