Case Study: Scoot achieves 85% higher accuracy and 50% reduced billing effort with AntWorks' Cognitive Machine Reading

Scoot Flies High with Automated Invoice Processing

Scoot, a subsidiary of Singapore Airlines, processes roughly 2,000 invoices a month—many 100–200 pages long—from over a thousand suppliers. Their template-based OCR and manual BPM entry struggled with highly variable formats, causing repetitive work, frequent intervention and growing processing time as the business expanded. Scoot engaged AntWorks to address the volume, variability and need for an AI solution that did not require in-house data science expertise.

AntWorks deployed Cognitive Machine Reading (CMR) to pick up tickets from Service Now, extract and categorize invoice data, flag exceptions for human QC, and route clean data to Service Now or Finesse System. The AntWorks CMR implementation delivered an 85% increase in accuracy, cut overall invoice processing time by 33% (with some invoices reduced from 3 hours to 30 minutes), halved the effort to create a bill in the financial system, and saved hundreds of hours on exception handling—enabling on-time payments and scalable financial operations.

Scoot