The Explorer's Guide to Cloud Native GenAI Platform Engineering

This conference talk provides a practical roadmap for building generative AI infrastructure from the ground up in cloud native environments. Navigate through three key phases of GenAI platform development, starting with a Thinnest Viable Platform (TVP) that demonstrates minimal setup requirements for LLM inference on Kubernetes. Progress to the MVP phase exploring essential platform components including LLM gateways, vector databases, and KServe deployments. Dive into advanced platform features such as intelligent load balancing for LLMs, observability patterns, and performance optimization techniques. Gain practical insights and lessons learned from real-world GenAI platform engineering as Max Körbächer from Liquid Reply and Alexa Griffith from Bloomberg share their expertise. Rather than focusing narrowly on isolated features, understand what a complete, functional GenAI platform actually looks like and how to build one using cloud native technologies.