UC2 · Step 4 of 5 — Apply the correction; downstream clears

Kavi Patel · Timekeeper · Tuesday, 14:11

Kavi clicks the teal Apply correction · Save button at the bottom right.

What happens, in order

1. The desktop app sends the resolution to the ASP.NET Core API with an optimistic concurrency token.
2. The API re-validates server-side — the trimmed activity (10:30 → 14:30) no longer exceeds the punch out (14:30), so the hard fail clears.
3. SQL Server records the new activity end time and an audit event ("K. Patel trimmed activity stop from 15:02 to 14:30 to match punch out · 2026-06-09 14:11").
4. The error disappears from Kavi's queue. The count of hard fails drops from 12 to 11.
5. The Downstream Check indicators update in real time:
   • 🟢 → 🟢 PLP feed: was blocked, now unblocked. Next poll (within 60s) will pick up the corrected activity.
   • 🟢 Piece-rate (job 0055): unit captured stays valid because the trim didn't affect unit count.
   • 🟢 Activity Errors report: this error is now off the report.
   • 🔵 Audit: new event recorded.
6. The Activity Errors report (the day-one report from the proposed solution) drops this line on its next refresh.
7. The next error in Kavi's queue auto-loads in the right pane. He keeps moving.

What Maria (staffing clerk) sees

Nothing changes in her workspace right now — she keyed the original assignment three hours ago and has moved on. If she opens the Activity Log for that shop order, she'll see Kavi's correction event in the history, attributed to K. Patel with a clear reason.

What Jaspreet (supervisor) sees

The hard-fail count on his Open Exceptions KPI tile drops from 3 to 2 on the next 60-second refresh. The line where T. Brown works no longer carries this error in the exception status panel.

What does not happen

• Kavi does not open Kronos and edit a punch.
• Kavi does not email anyone to say "I fixed it."
• Kavi does not wait for an end-of-day batch to refresh the downstream feed.
• Maria does not get pinged unless the trim is contested later (she can open the audit and see what happened).

Why it matters

The cost of an unresolved hard fail isn't just "Kavi has to fix it." It's the downstream consequences — PLP feed blocked, Activity Errors report carrying it forward, supervisor metrics off by 32 minutes, payroll close held up. The new flow makes "resolve the error" and "fix all the downstream consequences" one action, not five.

⬅ Prev: Step 3 — Choose a fix · Use Case Index · 🏠 Demo Home · ▶ Next: Step 5 — Recap