Cloud Foundry Day North America 2025

As AI adoption grows, organizations increasingly need efficient ways to serve AI models in cloud-native environments without excessive infrastructure complexity. While Cloud Foundry is not traditionally used for AI/ML workloads, it can still serve as a developer-friendly platform for deploying lightweight AI inference models.

This talk explores how pre-trained, pruned AI models can be efficiently deployed within Cloud Foundry, benefiting from its developer experience and streamlined application management while leveraging Kubernetes-backed Cloud Foundry deployments (cf-for-k8s) for scalability. We will discuss best practices for hosting AI inference services within Cloud Foundry, handling resource constraints, and ensuring efficient model serving without requiring full-scale ML platforms.

By the end of the session, developers and operators will understand how Cloud Foundry can be leveraged for AI-powered applications, enabling AI-driven services without the complexity of full-fledged ML infrastructure.