What was deliberately excluded from the numbers, and why.
Simulated data. Every capture in this demo is
origin: "simulated" — invented, not recorded on a jobsite. These
pages show how a productivity factor is derived. No model-accuracy
figure is reported anywhere: simulated captures may train a model and may never
measure one, so an accuracy number computed here would be meaningless.
1 labour-hours record(s) held back from the join — unmapped_cost_code:04-320. Surfaced rather than joined, so a bad cost code cannot silently become a productivity factor.
1 estimate(s) abstained — the model declined to guess rather than return a low-confidence number. An abstention is absent from the maths, not counted as zero installed quantity.
What this dashboard will not show
No individual-worker productivity view. Captures carry a captured_by field for provenance, and nothing groups by it. There is no table, column, or derived view that aggregates installed quantity per worker — the constraint sits in the schema, not just in the UI.
No accuracy claim from invented data. Accuracy is reported only against a held-out set of field and self-measured captures. Every capture here is simulated, so none of them may measure a model.
Assistant · fills forms and explains numbers. It never sets a quantity,
an abstention, or a face-blur declaration — those stay yours.