One truth for thirty offices
- 30+ institutions, nationwide
- 1 engineer, end to end
- 100% of changes audit-logged
The problem
Headquarters planned training quotas in a three-sheet Excel workbook full of formulas. Each institution kept its own copy. Numbers drifted, files conflicted, and nobody could say which version was true.
The moves
Start with the operators
I interviewed the people who maintained the workbook and mapped every formula to a rule the system could validate, so wrong numbers are rejected at entry instead of found at month end.
Dashboards people can read
Quota monitoring, an interactive map of every institution, and before/after audit logs on every change: who changed what, and what it was before.
A stack sized for the job
Go for a low-footprint API, PgBouncer for nationwide concurrent use, Redis for caching and queued Excel imports, and Garage for S3-compatible file storage.
The outcome
One real-time source of truth for training quotas and needs across the network.
Old Excel files can still be imported, so nobody had to retype history.
- SvelteKit
- Go
- PostgreSQL
- PgBouncer
- Redis
- Garage S3
- Docker
Note: The client is anonymised. No screenshots or data from the real system are shown.