inSTREAM – Automated Invoice Processing

Automatically capture and extract data, verify, validate and deliver structured data to Line of Business Systems.

The finance operation within a business, especially that of an Accounts Payable (AP) team carries a high level of responsibility to ensure invoices are correct, approved and paid on time. Yet the tasks carried out by the AP Clerks are time-consuming, repetitive and sometimes error prone.

With inSTREAM™’s Intelligent Document Processing, these tasks can be streamlined through AI, Machine Learning, OCR and RPA, delivering a simpler, more efficient invoice management system, improving processing time, compliance, supplier relationships and opportunities for early payment discounts.

Capture and Data Extraction

• Use UiPath robots to deliver inbound invoices to inSTREAM or deliver directly via post (paper), email, attachment, fax or EDI
• Extract key data – inSTREAM’s semantic engine and Natural Language Processing means there is no dependence on templates for document classification and data extraction

Validate, verify and enrich

• inSTREAM applies business rules to perform duplicate checks, invoice splitting, VAT calculations, PO matching, 2 - 3-way matching, verification of supplier details, currency and language conversion
• Machine Learning algorithms enable continual process automation and a ‘Human in the Loop’ document exceptions workflow

Delivery and review

• inSTREAM can be configured to ask a UiPath Robot to carry out further tasks for it by using the UiPath Add Queue ITEM REST API to create a job in the queue. These would typically be asking the robot to interact with a line of business system via the presentation layer to either lookup data for inSTREAM or enter data into the system.
• Dashboards and reports provide organisations with visibility of processes and people enabling monitoring of costs, quality, capacity and productivity leading to greater control. inSTREAM provides analytics and insights into audit trails, process insights and performance analytics.

Intelligent automatic data recognition, classification, extraction, validation and enrichment from unstructured, semi-structured and structured documents Semantic engine and Natural Language Processing mean there is no dependence on templates for document classification and data extraction Significantly improves efficiency through reducing AHT and errors Software-as-a-Service (SaaS), is fully managed and supported 24 x 7