Prerequisites¶
Tools and accounts you need before any quickstart.
Local tooling¶
- Java 21. Verify with
java -version. - Gradle 8.5+ (or use the project's
./gradlewwrapper). gcloudCLI for GCP authentication.
GCP setup¶
- A GCP project with the Vertex AI API enabled.
- Application Default Credentials set up — run
gcloud auth application-default loginonce. - A
europe-west2(London) oreurope-west1(Belgium) region selected for Vertex AI, depending on whether you need UK or EU residency.
Optional but recommended¶
- Cloud KMS with a CMEK in your chosen region — needed for the
dora,fca-sysc,pra-*,nhs-dspt,ehdsprofiles. - Confluent Cloud / Google-managed Kafka for the audit sink in production. Stdout is fine for development.
Quick verification¶
java -version
gcloud auth application-default print-access-token # should print a token
gcloud config get-value project # should print your project
If any of these fail, the quickstart will too. Fix here first.
What you do NOT need¶
- Anything from
langchain4j. Regulus' ADK path doesn't require it. - A pre-existing audit pipeline. Regulus ships one (Kafka or stdout).
- A pre-existing policy DSL. Regulus' profiles cover the common ground; custom policies are an opt-in.