Accurate data is the foundation for an efficient, effective Accounts Payable (AP) process.
Intelligent capture software that includes document imaging, separation and classification of invoices, and data extraction and validation introduces integrity in the downstream AP processes. Captured early in the process, invoices are sent in the electronic format to the appropriate employees for review, discrepancy resolution, and approval. This helps to ensure that invoices are approved within a week, with security and controls embedded. Intelligent capture eliminates the tiresome work of manual data entry and its associated errors, driving the return on investment in AP automation.
Evolution of Capture in AP
The first commercial applications of document imaging and capture software were pioneered over forty years ago. By integrating a scanner with optical character recognition (OCR) software, developers set out to “read” an image, identify its information and automate delivery of that information to other applications. The goal was to reduce the time, complexity, cost and errors of the manual process. While capture software was evolving considerably, the ability to achieve low-touch processing of invoices has proven elusive. Invoices present specific challenges, among them: There is no standard format for an invoice; every vendor delivers information in their own way. Many other types of documents arrive with invoices, including memos and credit applications. Information must be validated against other enterprise data, including vendor identification, employee identification, purchase order information, and receipt information.
The industry expectations today are that intelligent capture software will:
- Read all document-based information
- Separate and classify different documents that have been received
- Extract invoice information
- Validate invoice information
- Exchange information with enterprise systems
The first generation of capture software enabled the reading of all information. The following evolution provided the ability to focus on a subset of the information essential for the transaction. This was initially build around the concept of fixed forms, or structured documents where the same information is always located in the same area of the page. The ability to separate and classify was added.
The most elusive goals, the extraction and validation of invoice information, required two important developments: The software was enhanced to “learn” about semi-structured or unstructured documents. This was initially directed at an operator reviewing a scanned document, identifying the information to be extracted and its precise location, and creating a template for the document. This information was stored and applied when a similarly looking document was scanned. This labor-intensive effort was further hampered by the need to review and modify the template whenever a change condition was identified. Inevitably, the proliferation of templates resulted in the degradation of the performance of the application. The desired data quality proved difficult to achieve. This technology then evolved into “intelligent” capture — the ability to apply rules for finding the required data, tying it to surrounding keywords and other elements.
Instead of memorizing its fixed location, today’s technology automatically find the data among varied layouts following a logic that is similar to human reasoning. Intelligent capture software delivers more accurate, clean data, reduces errors, and helps achieve the lowest cost by reducing entry time. It delivers an attractive return on investment – typically within a year.
Keep reading. Download the full white paper:
Request More Info
