Building a Kubernetes Platform comes with its own set of challenges. Scaling across teams, securing multi-tenant workloads, and (how can we forget!), handling the surging demand for AI infrastructure.

Join us for a no-BS workshop where we move beyond the theory. Led by industry experts Daniele Polencic, Artem Lajko, and Saiyam Pathak, this event provides a practical 'blueprint' for the modern platform engineer. We will dive deep into creating seamless developer experiences, managing secure multi-tenancy, and leveraging virtual clusters to orchestrate GPU-heavy AI workloads without the overhead of physical cluster sprawl.

What will you learn:

• A hands-on workshop on how to build a production-ready platform that works not only in managed cloud Kubernetes, but also on edge, virtual clusters, and environments without managed Kubernetes.
• AI-Ready Infrastructure: Best practices for sharing and scaling GPU resources for LLM and ML workloads.
• Panel discussion and AMA: Bring your real problems — three platform engineers will diagnose and solve them live. Whether you're choosing an isolation model, debugging noisy neighbors, or figuring out how to roll out changes across 50 tenants, this is your chance to get answers. Come with questions; leave with solutions.

Agenda: 

11:30 -12:30pm: Arrival & Food

12:45 - 12:55pm: Kickoff - Daniele Polencic

1:00 - 2:30pm: Building a Multi-Tenancy Platform Beyond Kubernetes - Artem Lajko

2:30 - 2:45pm: Recharge break

2:45 - 3:45pm: Running LLM Workloads on Kubernetes - Saiyam Pathak

3:45 - 3:55: Recharge break

4:00 - 4:45pm: Architect’s round table and AMA with Jinhong Brejnholt, Artem Lajko, Saiyam Pathak, Daniele Polencic

4:45 - 5:30pm: Happy hour

‍