Platform Engineering Vienna

Join us at our March Meetup event for great talks, discussions and beers. Big thanks to Posedio for hosting!

Agenda:

• Doors open 5:30 p.m.
• Talk 1: 6:00 to 6:30 p.m.
• Speaker: Christian Kirchknopf
• Title: Building a Secure In-House LLMaaS Platform: BRZ's Approach to AI for Federal Government
• Abstract: In this talk, I demonstrate how we (BRZ) have developed an Enterprise-Ready LLM-as-a-Service platform that provides government applications simple and secure access to Large Language Models – without dependency on public cloud providers and without data sharing with third parties.
  Through a unified API, AI applications gain access to locally operated models (e.g., Mistral) or optionally to public cloud providers – depending on requirements, costs, and data protection needs of the workload.
  Discuss the architectural and product decisions: vLLM as a GPU-optimized inference engine, a central AI gateway that enforces multi-tenancy, security, and comprehensive observability.
  A brief live demo will show how applications and developers access AI models through a single API – with token-based quotas per tenant and intelligent routing.
  An experience report from an exciting AI infrastructure project in Austria.
• Break: 6:30 to 6:45 p.m.
• Talk 2: 6:45 to 7:30 p.m.
• Speaker: Damjan Gjurovski
• Title: What is A(I) platform?
• Abstract: In this talk I will discuss the need of platforms in the age of AI. I will outline some hard-won lessons from my previous experience in developer platforms, data platforms and SRE teams, and discuss how these apply to AI platforms.
  On the one hand, I will make the argument that AI makes certain aspects of platforms and platform engineering unnecessary, but on the other hand reinforces the need for other platform components.
  Finally, I will attempt to differentiate between the two new platform types out there – AI platforms and AI Engineering platforms and discuss the capabilities and requirements for each.
• Networking: start 7:45 p.m.