Decoding AP Automation
You have a pile of invoices to reconcile after having purchased goods and raw materials for your business. Small business owners like yourself know the importance of time and resources, and how critical it is to achieve maximum results with optimal usage of your resources.
Maximizing output from human and structural capital helps propel operations. And, subjecting your talented staff to mundane, uninspiring work is not one of them. Accounts payable is a labor-intensive process that is critical to the sound management of a company’s finances. What you need is a system that will remove the drudgery from invoice processing and have your employees attend to more growth-oriented tasks. AP automation is a good investment for companies looking to streamline payables and also achieve positive ROI in the long-run.
A typical AP Automation tool is preconfigured with process-specific intelligence to automatically acquire data and documents from any source including email, scan, MFPs, fax, etc. This multi-channel capture helps extend the application to any document-dependent process in any industry.
Apart from accounts payable, other typical application areas include sales order processing, medical claims management, check remittances, etc.
Acquired data is processed to make it ‘business-ready’ and applicable for use in ERPs and similar host systems.
Captured data is automatically classified and separated based on document type and defined business rules. In case of accounts payable, this means the application can distinguish between PO and non-PO documents, statements and BOLs, etc. In-built intelligence and program logic in the AP Automation software helps make these distinctions automatically. Data classification is an important step in document processing as it helps prepare and categorize data for the next step, which is to extract required details from documents.
When it comes to data extraction, most AP Automation vendors offer packages where the solution is pre-configured to capture header/footer details, line items, etc. as per your business requirements. OCR is the dominant method by which data from documents is extracted. But companies today have come up with more advanced techniques where Machine Learning is used along with OCR engines to automatically extract information from even low-quality images and poorly documented files. The type of extraction methods used depend on the quality of the source file and user requirements.
A good OCR engine should be able to support data extraction from files in many different languages.
Book a demo now
OCR is primarily a typewritten text recognition software. Popular OCR engines in the market include Nuance, AI Cognitive Services, Tesseract, etc. These do a good job of extracting text of a decent font and resolution. In case the text isn’t sharp enough to be recognized and extracted, an image preprocessing tool is used to clean up the text. Clean up could mean applying noise cancellation, removing blurring, adjusting for color and brightness, etc.
Additionally, accounts payable automation software should be familiar with different forms of text and be able to read handwritten text, signatures, and even watermarked information. ICR/IWR/OMR are other recognition tools that essentially recognize, extract, and convert data from paper forms to machine-readable form for further processing in order to make it ‘business-ready’. Because these tools commonly help with recognizing and extracting handwritten text, employing these along with OCR and Machine Learning increases the intelligent capture capabilities of AP Automation software manifold.
In spite of all these capabilities, there are chances that not all the required details will be captured. Latest features in AP Automation software come with easy-to-use capture technology where users can pull out the data that was ‘missed’ during the extraction stage through simple point-and-click (for selecting singular words or numbers) and rubber band (for selecting an entire row/column or table) techniques. These user actions are then remembered by the accounts payable software for the next time that it encounters a similar document— the accounts payable automation software emulates those user actions and performs the extraction automatically, reducing user intervention progressively.
Extracted data needs to be validated and verified before it is used in any business application. In the case of accounts payable operations, data validation involves performing 2-way (tallying PO and invoice details) and 3-way (tallying the contents of PO, receipt of goods, and invoice) matching, arithmetic checks, vendor lookups where data is validated against a master database in a corresponding integrated ERP or accounting system, etc. Data may be further verified by the user in case of any discrepancies before it is considered business-ready for use in any business application.
Considering the level of sophistication and inherent intelligence these accounts payable automation solutions come with these days, there is no reason why you should not deploy one for managing your payables. After all, a smart business is a successful business, and AP Automation makes businesses smart!
Ready for AP Automation? Contact Artsyl.
AP Automation & Shared Services
Managing Working Capital the Intelligent Way
Invoice Workflow Automation
Making the Most of Machine Learning and IPA
An Ultimate Guide to Invoice Processing
Automating Accounts Payable and Accounts Receivable