In this interview on theCUBE Research, host John Furrier sits down with Purnima Padmanabhan of Broadcom to discuss Tanzu Platform agent foundations, artificial intelligence-ready data foundations and secure open-source stewardship at VMware Explore 2026. Padmanabhan explains sandboxed agent runtimes and how these components integrate with VMware Private AI Cloud to accelerate secure agent development and deployment; they emphasize secure design and enterprise readiness. Padmanabhan states that Tanzu Platform provides a deny-by-default agent runtime with sandboxed containers, bind-based credentialing and AI-ready data foundations that enable curated role-based access control RBAC-controlled datasets. They note Broadcom is investing in Mythos scanning and Enterprise Spring patching to secure open-source dependencies and reduce vulnerability exposure. Padmanabhan highlights coding agents as a primary production use case and reports 2–3x productivity gains and a 70% reduction in time to find issues. This discussion provides practical insights for cloud native developers, DevOps teams, security operations and AI infrastructure professionals seeking secure agent deployment strategies, governance for open-source components and improved developer productivity.