Confidential AI, Self-Service: Rafay + Fortanix Confidential Containers Demo
Organizations want to apply AI to their most sensitive data: recorded customer calls, account numbers, personal details. Often the GPUs that run it belong to someone else, such as a neocloud or sovereign AI cloud. Confidential computing lets that work run in a hardware-protected environment. Rafay makes it a self-service offering that operators can deliver to every tenant.
This demo shows how Rafay runs confidential AI apps through its integration with Fortanix for CNCF Confidential Containers (CoCo). A bank's speech-to-text app transcribes customer calls inside a confidential container, so the data stays protected from the infrastructure operator while it is processed. Fortanix handles attestation and key management, and it releases keys only to an approved build. Rafay automates the platform underneath: clusters, blueprints, SKUs, and the self-service portal end users launch from.
The walkthrough follows four roles:
• Enterprise user: launches the confidential app on demand from Rafay Developer Hub
• Neocloud admin: onboards GPU infrastructure to Rafay and gives the bank its own tenant
• Application developer: encrypts the container image and registers the build in Fortanix Confidential Computing Manager
• Enterprise admin: builds a confidential cluster with Rafay blueprints and publishes the app as a SKU in SKU Studio
CloudAway and Apex Trust Bank are hypothetical companies created for this demo.
Learn more about Rafay: https://rafay.co
Learn more about Fortanix: https://www.fortanix.com

Trusted by leading enterprises, neoclouds and service providers

.png)







.png)







.png)

















