Automation · Regional logistics company
Taking the typing out of invoices
Over 500 invoices a month were being keyed in by hand across three departments. We automated it, cutting processing time by 80% and getting rid of data entry errors.
- Outcome
- 2 days → 3 hrsto close the books at month end
- Duration
- 10 weeks
- Tools used
- Python, OCR, PostgreSQL, React, REST APIs
The problem
The company processed over 500 invoices a month by hand across three departments. Staff spent hours every day typing in data, checking it against spreadsheets and chasing approvals by email. Mistakes were common, and month-end reconciliation regularly took an extra two days.
What we did
- 1Mapped the whole invoicing process with people from finance, operations and management
- 2Designed an automated process that reads each invoice, checks it against the purchase order and sends it to the right person for approval
- 3Built a simple dashboard showing where every invoice is and which ones need attention
- 4Set up automatic email reminders for approvals and escalations
- 5Ran the old and new processes side by side until the results matched, before switching over
What changed
- 80% less time spent processing invoices
- Data entry errors eliminated
- Month-end reconciliation down from 2 days to 3 hours
- Staff moved from data entry to more valuable work
- The project paid for itself within 4 months of going live
Dealing with something similar? Tell us about it.