LM Info optimizes invoice processing with Syslore Match
Syslore Match fuzzy matching system has enabled LM Information Delivery to decrease costs and manual work by significantly automating their invoice processing.
LM Information Delivery (LM Info) is the leading Northern European company focusing on subscription management, information delivery, electronic publications and database services. The company has been growing very fast during the last years and they currently have offices in Finland, Sweden, Norway, Denmark, Estonia, UK and Holland. As a result of this growth, LM Info needed an improved purchase invoice processing system. The goals of LM Info were clearly defined:
- Provide the best service for a growing amount of international customers and suppliers
- Increase and optimize their internal invoice processing efficiency
To achieve these goals, LM Info developed a system called LibInvoice. It is a centralized invoice workflow and purchase ledger system for processing all electronic and paper invoices throughout the organization. While the amount of electronic invoices is growing, paper invoice volumes are still significant at LM Info - especially from foreign publishers from multiple countries with several languages.
As part of the invoice scanning and data capture process, LibInvoice uses an OCR engine for automatic text recognition from scanned invoice images. However, the OCR recognition process was not performing well because invoices contain lots of inexact, incomplete and misspelled supplier and order information. To solve OCR recognition problems intelligently, LM Info uses Syslore Match fuzzy lookup software to match supplier, customer and order data against a reference database.
"We are very happy and satisfied with the Syslore Match solution”, says LM Info's software developer Mikko Kivistö. “Most importantly, our text recognition level jumped from 20% to 80%, greatly exceeding our expectations. Also, from the developer viewpoint, the solution is easy to deploy and the java-doc with its examples is impressive. There has not been any issues with the maintenance of the system, so I can warmly recommend it to other developers needing intelligent fuzzy matching capabilities".
More information about LM Information Delivery, please visit: http://www.lminfo.fi/