Building Secure MLOps Pipelines With KitOps + Argo Workflows

Learn to build secure and reproducible MLOps pipelines by integrating KitOps and Argo Workflows in this 27-minute conference talk from CNCF. Discover how KitOps, a CNCF Sandbox project, simplifies packaging machine learning models into immutable, attestable units with full traceability, while Argo Workflows provides Kubernetes-native orchestration for secure and reliable model deployment. Explore the challenges organizations face when scaling ML model deployment and understand how this integration streamlines the entire process from development to production. Walk through real-world pipeline examples that demonstrate secure model packaging and deployment in a reproducible manner, while learning to eliminate configuration drift through GitOps-driven enforcement. Gain insights into embedding security checks into ML-specific CI/CD processes and reducing friction between model creation and deployment. Master techniques for improving compliance through verifiable artifacts and understand how this integration benefits both ML engineers and DevOps teams. Acquire practical knowledge for building production-grade ML pipelines that maintain high security standards without compromising on efficiency or reliability.