PII Controls
llm.port can apply configurable PII handling controls before external model calls.
What this helps with
- Reducing sensitive data exposure risk
- Enforcing privacy behavior consistently across workloads
- Supporting regulated use cases
Policy model (public view)
PII behavior can be configured at system and tenant scope, with tenant-specific policy taking precedence when needed.
Common policy outcomes
- Redact sensitive values before outbound calls
- Allow or block requests based on policy
- Preserve operational observability for governance review
Deployment guidance
- Start with conservative defaults
- Validate policy behavior in staging
- Document exceptions and approval workflows
This public page intentionally omits internal implementation, storage, and pipeline internals.
Screenshots

