One AI API to Power Them All

Llama Stack unifies the core components of AI development into a single, cohesive API, eliminating the complexity of managing multiple disconnected tools. In this session, we’ll show how Llama Stack provides a standardized approach to inference, RAG, agent workflows, tool integration, safety measures, evaluations, and telemetry—all within a flexible plugin-based architecture. Through live demos, we’ll demonstrate how developers can build and deploy AI applications seamlessly across different environments, from local machines to cloud and on-prem infrastructure. We’ll also explore how Llama Stack integrates with Kubernetes and OpenShift, providing a scalable, enterprise-grade foundation for running AI workloads efficiently. Whether using the CLI, SDKs for Python and TypeScript, or mobile interfaces for iOS and Android, Llama Stack enables developers to build production-ready AI applications with minimal overhead.