PDF.co
Document ProcessingGive agents full PDF processing power. Convert, merge, split, extract, classify, compress, secure, and parse invoices with AI. 17 actions covering every document operation agents need to handle unstructured content.
What This Integration Enables
PDF.co integration sits in the document processing layer of agent workflows. Before data can flow into your ERP, CRM, or accounting system, unstructured documents need to be processed and normalized. PDF.co handles that transformation: extracting text, parsing invoices, converting formats, and generating documents from data. It is the processing layer between raw documents and operational systems.
Without FlowRunner
With FlowRunner
Use Case Scenarios
AI-Powered Invoice Processing Pipeline
A vendor submits an invoice as an email attachment. Parseur extracts the raw document. The agent passes it to Parse Invoice with AI, which extracts vendor name, invoice number, line items, amounts, and due date. The structured data is validated against the vendor record in Acumatica. If it passes, the agent creates the bill and attaches the original PDF. If extraction confidence is low, the agent routes to the AP team with the raw document and extracted fields for review.
Contract Package Assembly
A deal closes and requires a compliance package: signed MSA, SOW, DPA, and order form. The agent retrieves each signed document from DocuSign, converts any Word documents to PDF using Convert Multiple Formats to PDF, merges all four into a single package using Merge PDF Documents, applies security using Add Security to PDF, and stores the protected package in S3. The complete signed package is assembled and secured without anyone opening a PDF editor.
Document Classification and Routing
An inbox receives mixed document types: invoices, PO confirmations, delivery receipts, and vendor correspondence. The agent retrieves each document and passes it to Classify PDF. Based on the classification, each document routes to the appropriate processing workflow: invoices to AP, POs to procurement tracking, delivery receipts to inventory management. Documents that cannot be classified route to the ops team for manual classification.
Human-in-Loop Highlight
Parse Invoice with AI returns confidence scores for extracted fields. When the confidence for a key field (total amount, vendor name, invoice number) falls below a defined threshold, the agent does not push low-confidence data into the ERP. It pauses and routes to the AP team via Slack: "Invoice extraction from [filename] returned low confidence on the total amount field (extracted: $4,200, confidence: 61%). Review the document and confirm the correct amount before I create the bill." The team verifies. The agent proceeds with confirmed data. The ERP gets reliable inputs.
Agent Capabilities
17 actionsContent Extraction
5- Convert PDF to Text (Basic) Extracts plain text content from a PDF document. Used for lightweight text extraction where structure is less important than raw content.
- Convert PDF to Text (Advanced) Performs more sophisticated text extraction with better handling of complex layouts, tables, and multi-column documents. Used for financial documents, forms, and structured reports where layout matters.
- Classify PDF Uses AI to classify a PDF document into a category. Used in document routing workflows: the agent classifies an incoming document as invoice, purchase order, contract, or report, and routes it to the appropriate processing flow.
- Parse Invoice with AI Uses artificial intelligence to extract structured data from an invoice: vendor, invoice number, line items, amounts, dates, payment terms. Used as the primary extraction step in AP automation workflows, especially for invoices from vendors that do not follow a consistent format.
- Read Barcodes Detects and reads barcode values from a PDF or image. Used in warehouse and logistics workflows where barcoded documents need digital processing.
Conversion and Export
3- Convert Multiple Formats to PDF Converts Word, Excel, PowerPoint, HTML, and other formats into PDF. Used when agents need to standardize document format before storage, signing, or delivery.
- Convert PDF to Images Converts PDF pages into image files. Used when document content needs to be displayed as images or when downstream systems accept images but not PDFs.
- Convert PDF to Multiple Formats Converts a PDF to Word, Excel, CSV, or other formats. Used when extracted data needs to be in an editable format for further processing.
Document Operations
6- Merge PDF Documents Combines multiple PDF files into a single document. Used to assemble proposal packages, combine multi-page invoices, or create consolidated reports.
- Merge Documents (Multi-Type) Merges documents of different types (PDFs, Word, images) into a single PDF. Used when evidence packages or proposal bundles contain mixed source formats.
- Split PDF Divides a PDF into separate files by page range or structure. Used when a combined document batch needs to be separated for individual processing.
- Delete PDF Pages Removes specific pages from a PDF. Used in document preparation workflows when a combined file contains pages that should not be included in a delivered document.
- Compress PDF Reduces the file size of a PDF. Used when documents need to meet size limits for email delivery or storage quotas.
- Generate Barcode Image Creates a barcode image. Used in inventory and logistics workflows where documents need tracking barcodes.
Security
2- Add Security to PDF Applies password protection and permissions to a PDF. Used when sensitive documents need to be protected before delivery.
- Remove Security from PDF Removes password protection from a PDF. Used when a protected document needs to be processed by an agent that has the authorized password.
Jobs
1- Check Background Job Status Retrieves the status of an asynchronous PDF.co processing job. Used when large documents are processed asynchronously and the agent needs to wait for completion before proceeding.
Start building with PDF.co
$100 in credits. No card required. Connect in minutes.