From “Need AI” to a Shippable Service: Platform Engineering Workflow @VMwareTanzu
From “Need AI” to a Shippable Service: Platform Engineering Workflow  @VMwareTanzu
Uploaded December 2025 | Updated September 2026, 54 minutes ago
“I need an AI model” is a request - not a spec, not a service, and not an operating model. This episode shows how platform engineers translate that ambiguity into standardized, self-serve offerings via a service marketplace. We cover t‑shirt sized plans, working with AI/domain teams, and automating the approvals and guardrails so devs can bind AI capabilities to apps without a ticket chain.

Here's how to get AI setup in your platform 👉 https://TryTanzu.ai.

Tanzu Catsup is a weekly conversation about platform engineering, cloud-native operations, and building software in large organizations. We follow the work wherever it actually leads.

Check us out Tuesdays at 10am US Eastern/4pm Amsterdam time!
From “Need AI” to a Shippable Service: Platform Engineering WorkflowYou already have a platform. Your developers built it.If it aint broke, dont fix itTransitioning to Tanzu Platform 10: Plan Your Upgrade from Tanzu Application ServiceAccelerate Developer Collaboration: Service Publishing Demo in Tanzu Platform 10.3Top 5 Features We Love in Tanzu Platform 10.2How Tanzu Platform works with Kubernetes, VKS and Cloud FoundryYour supply chain scanner has a supply chain problemSixty percent of the Spring downloads were for project versions that were no longer supported!Why DIY platforms create more headcountAI is too easily distractedEnterprise AI is about app dev! 👨🏿‍💻🧑🏼‍💻No time for platform building and DIY hairballs⤵️
VMware Tanzu |

From “Need AI” to a Shippable Service: Platform Engineering Workflow

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER