Client Profile.
A leading information management software and service solutions provider, the client helps more than 400 global businesses to scale up their business processes. They leverage automation and technology to create, process, manage and distribute vast quantities and types of information including mail, invoices, drawings, web pages, financial reports and many other forms of electronic information.
Business Need.
The company’s data capture and processing solutions for financial documents required ongoing scanning of huge volumes of data to gather information, entering the data and matching invoice data against supplier payments. In order to speed up the process and improve the accuracy of invoice data processing, the company was looking for a partner with extensive experience in offering automated data management solutions.
Challenges.
- Manually extracting information like supplier number and name, invoice number, currency, tax amount, net amount, purchase order number and invoice date was a cumbersome task.
- Disparate invoice formats from hundreds of suppliers amplified the need of skilled resources and advanced tools.
- Existing process was time- and labor-intensive extracting each data point accurately was critical.
- Invoices were often documented improperly, creating a huge backlog of unpaid invoices.
- Absence of a central system to manage critical accounts payable process.
Solution.
The entire invoice processing activities, from downloading scanned invoice images to extracting, validating and updating invoice details in the client’s web based KOFAX tool was automated. Scripts and bots backed with human intelligence were used to automate the entire invoice validation process. Macros were leveraged to generate a unique identification number which ensured timely payment processing.
The results:
- Huge volumes of invoices were processed monthly
- The average handling time per invoice was reduced up to 50%
- 100% accuracy & 100% compliance
- Set up a system to provide a detailed daily invoice status report, deviation reports on payments, etc.
- Reduced the volume of paper invoices and improved accuracy and control
- Client was able to access dashboards to track invoices.
Approach.
Receiving input documents from client
- The data team received an email alert, with the count of invoices, whenever the client placed scanned images of invoices in PDF form on the FTP server.
Classification of invoices
- Scanned invoices were classified according to vendor type, vendor name, invoice number, invoice date, amount etc.
- Macros ensured 100% accuracy in vendor-wise data segregation.
- Automated bots and custom macros routed invoices to different folders based on their approvers.
- Scripts and macros reconciled incoming invoices with purchase orders. Any mismatch was conveyed to the client.
- Bots generated exceptions for human intervention if the purchase order number was missing or if the invoice was not standard.
- The bot also sent a detailed report of processed and unprocessed invoices via email at the end of each day.
Data Entry of extracted information
- Invoices classified as valid/invalid based on checks against pre-defined parameters, were then moved for OCR data extraction; and the rest were held back.
- Data extracted from invoices was then sent for structured data entry in the web-based tool.
- Leveraging automated bots backed with human intelligence; the invoices data was then routed to approvers, according to their designation.
Quality check of data entered
- Using predefined rules for the quality check process ensured 99.5% data accuracy.
- Macros created alerts notifications for any discrepancy or error in data entry.
- Errors identified were corrected manually and automated.
- Rules/algorithms were modified based on what we learned.
