Replay
A controlled experiment that re-runs a captured workload shape against candidate execution profiles to compare cost, latency, and capture.
Replay is how BYOC and routing changes are justified before they ship. A replay run pins inputs, alias releases, and workload shape, then compares a baseline against a candidate and emits a signed report.
It turns "this should be cheaper" into "this was 18% cheaper on the same traffic," which is the bar for moving a lane.
Keep reading.
BYOC (bring your own cloud)
Running the inference data plane inside the customer’s cloud for dedicated SLOs, isolation, and explicit KV orchestration.
Alias release
An immutable, versioned snapshot of an alias’s resolution policy, so a routing decision can be reproduced exactly.
Evidence level
A label for how trustworthy a reuse measurement is, from provider_reported down to trace_estimated and unknown.
See it in practice.
Definitions are useful; measurement is better. Run a diagnostic on your own workload.