Fictional sample. Halvorsen Builders, its people, projects, clients, and pursuits are invented.
From: Linda Marsh, VP Operations Sent: Monday, October 5, 2026, 6:52 AM Subject: Staffing meeting Thursday
Can you get the staffing picture ready for Thursday’s ops meeting? Exports attached. Same questions as always: who’s slammed, who’s open, and where we have holes. Specifically:
- We still don’t have a super for Wash Park. GMP is December 15, and they start digging in January.
- Tony retires March 31 and Arvada isn’t done until May.
- Riverside interviews on the 20th. If we win it, who runs it? The owner will ask about ICRA.
- Union Station slipped again (see the 9/28 schedule update). I don’t think the sheet reflects that.
Don’t move anyone in the sheet. I want options I can put in front of the PXs, not decisions.
Attached files
| File | What it is | Quirks to expect |
|---|---|---|
| input-staffing-sheet.csv | The “Field + PM” tab of Staffing 2026-27.xlsx: one row per person, one column per month | Bare job codes mean 100%; ? means pencilled for a pursuit; St V and Union Stn are the same jobs as STV and UNI; Frank’s cells list jobs with no percentages; a “TBD Super” row at the bottom; free-text notes |
| input-projects.csv | Ops project list with phase dates and each job’s planned team | Union Station’s substantial completion moved from April to July 2027; planned teams use shorthand like AS 100% (ICRA) |
| input-pursuits.csv | CRM pipeline export | Win probability is a percentage string; Riverside’s start moved from February to March; staffing needs use the same shorthand |
| input-roster.csv | HR roster | Certifications, sector experience, travel, the new hire’s start date, Tony’s retirement, and Megan’s parental leave |
The agent normalized these into staffing-data.json, recording each ambiguity it resolved as an open question with its source row, then ran the bundled script to produce the brief and the interactive plan.