MuleSoft Automation: Invoice Processing Demo

Описание: It takes between 4.1 and 16.3 days for companies to process an invoice from receipt through payment approval, according to a research report by technology services company Aberdeen Group.

This is because employees in the Accounts payable department need to handle the manual effort still involved with processing invoices on a day-to-day basis.

Invoice processing is a critical finance process, and many companies are doubtful regarding the capabilities and success rates of RPA bots. Adding to that is the sheer number of Invoice formats that go into the Company's database and deal with the discrepancies. These challenges make it difficult to handle invoices automatically with a one-size fits all solution.

However, at Infomentum, we design the right solution. It automates a significant amount of manual effort required by Invoice processing. The solution can be a truly Human-Robot collaborative process.

In simple terms, key tasks, like monitoring shared drives/emails/folders for invoices and extraction of data using OCR, can be executed by a bot, while other tasks, like handling the anomalies and approvals of invoices, can be taken care of by the employees. Hence, the efficiencies in Invoice automation are increased by 70% (including automated success and human-in-the-loop success). Furthermore, the AI-based solution only becomes better with constant use, learning from the previous transactions.

In this example demonstrated in the video, the bot is designed to monitor emails with Invoice attachments to extract PDF files using Amazon Textract, the best-of-breed Intelligent Document processing engine. The bot automatically enters the extracted data into the Xero finance application. In case of data extracted below a set confidence threshold, the bot automatically routes the information to its human counterpart through tasks.

We tailor the bot as per customers' requirements and reuse the Invoice processing code and APIs to connect to Source applications and Finance systems. We designed the bot to handle other activities like vendor information matching, three-way matching, approval flow based on Invoice amounts, etc., to improve overall productivity and operational efficiencies.