Skip to main content
Request Demo
Browse help topics

Help and support

The Processing Dashboard serves as the central operational cockpit where users monitor, review, track, and manage the end-to-end lifecycle of documents being processed by Artificio's AI Agents.

Key Functionalities of the Processing Dashboard

The Processing Dashboard is divided into core operational pillars to ensure seamless Straight-Through Processing (STP) or structured Human-in-the-Loop (HITL) exception handling:

  • Real-Time Intake & Ingestion Monitoring: Visualizes documents flowing into the platform from various multi-channel streams (such as email attachments, FTP/SFTP folders, web portal uploads, cloud storage, or third-party system connections).

  • Queue & Status Management: Organizes documents dynamically into queues based on their real-time state:

    • Received / In Queue: Documents awaiting initial optical character recognition (OCR) and document preprocessing.

    • Processing: AI models are actively classifying the documents, categorizing data, and executing Key-Value Pair (KVP) and Table-Line Item extraction.

    • Needs Review / Exception: Documents flagged due to low confidence scores from the AI, structural anomalies, or failed business validation rules.

    • Processed / Exported: Successfully parsed and verified data pushed to downstream destination systems (such as ERP or accounting ledgers).

  • Data Verification & Exception Resolution Workspace: A dual-pane interface provided for documents requiring human intervention. On one side, the user views the original document image or PDF; on the other, an interactive form displays the extracted metadata fields (such as invoice numbers, totals, or utility line items) where low-confidence fields are highlighted for quick human approval or adjustment.

  • Anomalies & Validation Rules Engine: Displays visual alerts for calculation mismatches (e.g., table line items not adding up to the stated total amount), duplicate invoices, or unrecognized vendors, preventing erroneous data from moving forward.

Core System Workflow Driving the Dashboard

Behind the user interface, the dashboard relies on a series of connected background operations and AI agents to function seamlessly:

  • Ingestion and Image Cleanup: When a file is uploaded or pulled into the queue, the system triggers the Ingestion Agent and Preprocessing Agent. This step handles image cleanup, including deskewing, denoising, binarization, and language identification to optimize the text for reading.

  • Analytical Matrix Updates: The main grid views, charts, and key metrics (such as daily volume processed, error rates, and average handling time) rely on background tracking services that continually update the frontend analytics panels.

  • Document Intelligence & Machine Learning Feedback: When an operator views an exception, the system displays structured data payloads alongside precise bounding-box coordinates to highlight text locations on the screen. When a human adjusts data or clicks "Approve," the platform feeds this corrected information back into the Exception Resolution Agent, allowing the underlying AI to learn and improve its accuracy over time.

  • Workflow Integration & Final Export: Once a document clears the validation logic or is manually approved, the system's ERP Integration Agent maps the finalized data directly into the user's destination systems, automatically clearing the item from the active Processing Dashboard.