# DocSumo MCP for AI Agents AI Agent Connect > DocSumo MCP automates document data extraction and IDP workflows. It lets your AI agent pull structured data from invoices, bank statements, and ID cards while monitoring pipeline health and audit logs. You can check extraction confidence, identify failed jobs, and retrieve verified fields using natural language. ## Overview - **Category:** artificial-intelligence - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_DarJaMV6QupFu1Axj4v77AXHtRZpXKyo2RrpnsLx/ai-agent-connect - **Tags:** intelligent-document-processing, data-extraction, idp, invoice-processing, automated-pipelines, data-audit ## Description DocSumo lets you turn piles of messy documents into clean, usable data. If your team spends hours manually typing out invoice numbers or bank statement totals, this Connector changes that. You can connect your AI agent to the DocSumo platform to pull structured fields automatically. Instead of clicking through folders, you can just ask your agent to find a specific document by filename or list all the ones that successfully parsed. It also handles the oversight part of the job. You can check your account limits, see which files failed to process, and get a high-level summary of your pipeline health. It's a way to move your Intelligent Document Processing into a conversational workflow. You'll find it in the Vinkius catalog as a reliable way to bridge the gap between raw files and your internal records without the manual grind. ## Tools ### get_docsumo_account_metadata Check your account's usage limits and metadata. This helps you keep track of your platform limits. ### list_failed_doc_extractions See which documents failed the extraction process. This lets you identify and fix errors quickly. ### get_document_extraction_data Get the structured data from a specific document. This is the primary way to pull fields into your system. ### quick_idp_health_audit Get a high-level summary of your processing activity and success rates. Use this to monitor your overall pipeline. ### list_successfully_parsed_docs Find documents that were successfully processed and verified. This helps you confirm work is complete. ### list_docsumo_document_types See all your configured document types like invoices or bank statements. This shows your current setup. ### list_processed_documents List all your processed documents with an optional type filter. This lets you browse through finished work. ### list_latest_extraction_results See the most recent documents processed across all categories. Use this for a quick pulse check on recent activity. ### list_documents_awaiting_review Find documents that need manual review due to low confidence. This keeps you focused on only the files that need help. ### search_documents_by_filename Search for processed documents using a specific filename keyword. This is the fastest way to find a specific record. ## Prompt Examples **Prompt:** ``` List all processed bank statements. ``` **Response:** ``` I've found 12 processed bank statements, including 'Chase_June_2023.pdf' and 'Wells_Fargo_Q2.pdf'. | Document Name | Status | Date Processed | | :--- | :--- | :--- | | Chase_June_2023.pdf | Verified | 2023-06-15 | | Wells_Fargo_Q2.pdf | Verified | 2023-06-16 | Would you like to see the extracted summary for the most recent one? ``` **Prompt:** ``` Show me the extraction results for document 'DOC-5544'. ``` **Response:** ``` I've retrieved the data for 'DOC-5544'. **Extracted Fields:** * **Invoice Number:** INV-789 * **Vendor:** Cloud Services Inc * **Grand Total:** $4,500.00 * **Confidence Score:** 98% Should I list the individual line items? ``` **Prompt:** ``` Are there any documents that failed to process today? ``` **Response:** ``` I've checked the pipeline and found 2 documents with processing errors: * **Unreadable_Scan.pdf**: Image quality too low. * **Wrong_Format.jpg**: File type not supported. Would you like the detailed error messages for these items? ``` ## Capabilities ### Pull structured data from documents Get specific fields like totals, dates, or vendor names from files automatically. ### Identify failed extractions See which documents didn't process correctly so you can fix them quickly. ### Monitor pipeline health Get a summary of success rates and activity across your entire operation. ### List document types View all your configured categories like invoices, receipts, or IDs. ### Find documents by name Locate specific records instantly using a filename keyword. ### Flag manual reviews See files that need a human eye because the AI confidence was too low. ## Use Cases ### High-volume invoice processing A finance clerk asks the agent to pull all invoice totals from a batch of 500 PDFs. The agent uses get_document_extraction_data to build a summary table. ### ID card compliance auditing A compliance officer asks if any ID cards failed verification today. The agent uses list_failed_doc_extractions to provide a list of errors. ### Bank statement reconciliation An accountant asks for the latest bank statements. The agent uses list_latest_extraction_results to show the most recent files processed. ### Pipeline health monitoring An ops lead asks for a status update. The agent uses quick_idp_health_audit to report on the success rates of the current extraction jobs. ## Benefits - Stop manual data entry by using get_document_extraction_data to pull fields directly into your system without typing. - Spot errors faster with list_failed_doc_extractions to see exactly where a process broke down and fix it immediately. - Keep your team focused by using list_documents_awaiting_review to see only the files that actually need a human eye. - Monitor your entire operation with quick_idp_health_audit to see real-time success rates and processing activity. - Organize your workflow by using list_docsumo_document_types to see every category you have configured in the platform. - Find specific records instantly with search_documents_by_filename instead of digging through folders or spreadsheets. ## How It Works The bottom line is this Connector turns your AI agent into a high-speed data extraction clerk that handles the boring parts of document processing. 1. Add your DocSumo API key to your AI client settings. 2. Connect the DocSumo MCP to your agent. 3. Ask your agent to pull data, check status, or audit your workflows. ## Frequently Asked Questions **How does DocSumo MCP help with invoice processing?** It lets your AI agent pull specific data like totals and vendor names directly from your invoices. This eliminates the need to manually type that information into your accounting software. **Can DocSumo MCP handle bank statements?** Yes, it can extract structured data from bank statements. Your agent can pull these details into a summary or a database for easier reconciliation. **How do I find documents that failed to process?** The Connector allows your agent to identify which documents failed during the extraction process. You can ask your agent to list these errors so you can address them quickly. **Can DocSumo MCP audit my ID card processing?** Yes, it can help you audit processed ID cards. You can ask your agent to check the verification status or find documents that require a manual review. **How do I see my pipeline health with DocSumo MCP?** You can ask your agent for a high-level summary of your document processing activity. It will report on success rates and overall pipeline health instantly. **Can I search for specific files using DocSumo MCP?** Yes, you can ask your agent to find a specific document using a filename keyword. This helps you locate records in your history without searching through folders. **How do I get a DocSumo API Key?** Log in to your DocSumo account, navigate to the API section in your settings, and you can retrieve your unique API Key from there. API access is typically enabled for most plans. **What happens if extraction confidence is low?** DocSumo flags documents with low confidence for manual review. You can use the list_documents_awaiting_review tool to identify these documents directly via the agent. **Does the integration support custom document types?** Yes, as long as you have configured the document types in your DocSumo account, the agent can list them and retrieve extracted data for any of them.