Behind the artifactsA Day in My Life

What Wave 3, week 14 actually looked like.

The phase pages show you the artifacts. This is the day those artifacts came out of — a real Tuesday from the middle of Execution, the same week as the RAG dashboard and the 6% variance catch you've already seen.

7:30
AM
Coffee, then the overnight reconciliation run
  • First thing, before standup — check the dual-run reconciliation results from overnight
  • This is where I saw the revenue-sum flag for the first time: Snowflake running 6% high against source — no root cause yet, just a number that shouldn't exist
8:00
AM
Data Engineering standup
  • Fifteen minutes, stand-up format, no chairs — I raise the variance immediately, not left sitting in my notebook until the RAID review
  • Team's working theory is a load bug; I ask them to check the join grain before assuming that
9:00
AM
RAID triage & blocker clearing
  • Thirty minutes I protect every day — no meetings scheduled over it
  • Log the variance in the RAID as an open issue
  • Move the Wave 3 pipeline performance flag from "monitoring" to "at risk" — the thread that becomes the Wave 3 decoupling decision two weeks later
10:30
AM
Working session with the Sales Finance SME
  • This is the meeting that produces D-07 — I bring the grain-mismatch theory, she confirms the source carries one row per opportunity per daily snapshot
  • Real question isn't technical: does "close amount" mean booked or forecasted revenue? That's her call, not mine — I make sure everyone in the room knows that going in
12:00
PM
Lunch at my desk, building the Steering deck
  • Bi-weekly steering is Thursday — I use the quiet hour to turn this week's RAG status into the two-minute version
  • Data Engineering green, BI amber on the Revenue by Segment slip, and a new line for the variance — already ruled on before it reaches the deck
1:30
PM
Design authority session
  • Ad hoc forum I chair whenever there's a model decision that can't wait for steering — today it's D-07, formally
  • Log the ruling, note that it resolves the variance, and make sure the fix routes back to the transformation logic, not a one-off patch in someone's local branch
3:00
PM
Walking the floor with QA
  • I go sit with the QA lead instead of pulling them into a meeting
  • Reconciliation suite output together — row-count parity clean, sum parity now explainable by D-07 once the fix lands
  • This is where I hear, informally, that the Wave 3 pipeline will miss its performance target — worth more than the status report that confirms it two days later
4:30
PM
Sync with the Change Lead
  • Champions network check-in — Sales' champion recruitment is behind
  • Flag it as a risk to watch, not yet an issue, and agree a recovery plan before it becomes one
  • This is the "day one, not day-of" change management I keep pointing back to
5:30
PM
Closing the loop before I leave
  • Update the decision log with D-07
  • Move the Wave 3 pipeline risk from "monitoring" to "at risk," with an owner named
  • Write two lines in my own notes about the Product Director conversation I know is coming once that pipeline slip is visible to everyone else — tomorrow's problem, but I want to walk in already knowing my answer

Why this page exists

Every artifact in this case study — the RAID log, the decision log, the RAG dashboard — is the output of a day that looked roughly like this one. None of it happens in a document; it happens in standups, hallway conversations, and one working session with the right SME at the right time. The paperwork is just where I wrote it down afterward.