For the next two weeks this system is your workplace. A member's procedure needs approval; a score decides its path; a clock starts either way. Here's the whole machine — and what you personally will do to it.
Hold that trade-off. It's the same one you'll manage all course long with AI itself: automation above a confidence line, human judgment below it, and everything depending on where the line sits and who verified the scorer.
Written like a real team wrote it — mixed styles, old TODOs, a git history with moods. Twenty-four planted artifacts hide in there, and at least eight of them will be visited by more than one role. Your role sees its lens; the cohort sees the whole.
A generated-then-mutation-proven suite covering the auth flow, and the review standard to sign off any agent's tests.
NeoLoad scenarios and TruClient journeys that certify the SLA — then re-certify it at 24h in the capstone.
That 2.9s p95 cut to milliseconds with defended EXPLAIN evidence, and migrations that survive the rollback drill.
A spec and ADR set an agent can build from, governance that runs in CI, and the exec brief that wins the room.
Prior-auth agentic build and review fluency — then the bridge hands you to the full 10x Toolkit.
Baseline Week starts tomorrow: you'll do your role's tasks with no AI at all, on the clock. At the end of your pack you'll face the same tasks again, with everything you've learned. The delta — time, defects, rework — is measured, not promised. That's your evidence, in your numbers.