An authorization evaluator, versus a policy control plane.
OPA and Cedar do one job: they evaluate a decision from the inputs you give them, statelessly and fast. Governance on top of one needs everything that makes a decision safe to change and possible to prove, and you build all of it yourself.
What OPA is
A graduated CNCF project with a large community and its own policy language, used for authorization across Kubernetes and infrastructure. It stays in your cluster, and Swiftward is the control plane for the decisions that need memory and a record.
What Swiftward does differently
OPA answers "given these facts, is this allowed." It does not remember. The moment your policy needs "is this the fifth transfer this hour" or "has this user's trust score dropped," you add a database next to OPA and take on keeping those counts correct under load. Swiftward is stateful by design: counters, rate limits, sliding windows, and cooldowns are built in, ACID, and an event processed twice changes them only once.
What you build on top of OPA
Each of these is its own project:
- versioning, and rollback by naming the previous version;
- shadow mode and A/B on live traffic;
- a record that names the policy version behind each past decision;
- a dead-letter queue;
- human review that survives a restart;
- an audit trail an auditor accepts.
Together they are a platform that your team would build and then maintain. Swiftward is that platform, already built and maintained.