See it work. No sign-up.
Drop a few day-book rows in, watch it draft a clean payroll-ready file, then a flagged row where it isn't sure. That flag is the point: you always see what it's unsure about.
Your sample day book
One row per worker/day, columns: name, date (dd/mm/yyyy), hours, job ref.
Draft — confirms to me first
Waiting for your input…
This demo runs the actual DayBook pipeline — the same normalisation and confidence-scoring your data would go through in production.