This document links out to codelabs that demonstrate building specific Confidential Space environments from beginning to end. You can use them to familiarize yourself with how to set up Confidential Space environments, and adapt them to your own use cases.
-
Confidential LLM inference
Communicate with an LLM served in a Trusted Execution Environment by using the Prompt Encryption SDK.
-
Use Confidential Space with resources hosted outside Google Cloud
Request a self-contained token from Google Cloud Attestation that a relying party can verify offline.
-
Transact digital assets with multi-party computation
Perform multi-party computed blockchain signing using Confidential Space.
-
Share proprietary machine learning models in Confidential Space
Run a proprietary model in Confidential Space, which another party can use private data with.
-
Deploy a Confidential Space workload with MIGs using autoscaling, autohealing, and image updates
Automate your Confidential Space image lifecycle.
-
Use container image signatures to validate workloads
Sign your container image to avoid changing your attestation policy whenever a workload image updates.
-
Determine common data without compromising privacy
Walk through a scenario where two banks determine common customers without sharing account details with each other.