2026 Week 4
- Improved email-to-PDF conversion for emails with large images that don’t fit on a PDF page.
2025 Week 53
- Improved identity document cut-outs:
- Reduced unnecessary white space around cut-out
- Passport backsides are cut out even when the BSN is not visible
2025 Week 51
- Added the ability to use one workflow as an action within another and retrieve its output, providing greater flexibility in workflow design.
2025 Week 48
- Added support for Dutch (NL) driver’s licenses issued after November 2014:
- Implemented data extraction for Name, First Name, Date of Birth, Place of Birth, Date of Issue, Expiry Date, Issuing Authority, Document Number, and MRZ data from the front and BSN from the back.
- Implemented detection and cut-out for both front and back sides.
2025 Week 44
- Improved extraction of Pension type and Name from Pension Overview documents, to ensure more accurate categorization of Pension Type and correct extraction of full Pension Name, even when spanning multiple rows.
- Improved handling of different Mortgage Deed formats to ensure consistent and more accurate extraction of Mortgage Rank and Rights Holder values.
- Improved handling of superscript numbers in the Declaration of Work document to ensure they are no longer incorrectly attached to Income Type values, preventing the addition of unexpected characters (e.g., ', (, !) in the extracted data.
- Improved extraction of Loan Part data from Mortgage Deed Final Offer documents with the following updates:
- Updated duplication filter logic to ensure complete extraction of Loan Part.
- Added support for 10-year-old document layouts to enhance compatibility and accuracy.
- Improved handling of Interest Rate to ensure more accurate extraction when they span multiple pages.
- Improved extraction of Credit Amount from BKR documents, ensuring more accurate recognition for formats with decimal separators (e.g., '1.234,56') and whole numbers without separators (e.g., '1234').
2025 Week 42
- Improved ID cut-out accuracy when ID documents are placed near text, ensuring more reliable detection and extraction.
2025 Week 41
- Improved passport back-page cut-out functionality to support passports issued between 2014 and 2021.
2025 Week 40
- Updated ID detector to also find driver's licenses, residence permits and passport page backs.
- Improved signature detection for the Declaration of Work document type.
2025 Week 38
- Added extraction of the Employer Signature data point for the Declaration of Work document type.
- Improved email-to-PDF conversion by skipping unsupported attachments (for cases using the 'ExcludeAttachment' setting only).
2025 Week 33
- Improved extraction of Mortgage Rank and Rights Holder data from Mortgage Deeds, ensuring better handling of line breaks and dashes.
- Enhanced detection of Object Purpose and Renovated Value in Taxation Reports by incorporating additional keywords for more accurate data extraction.
- Improved support for older-format documents, including:
- Extraction of Taxation Date and Collateral Value from WOZ value reports.
- Extraction of End Date, Interest Rate, and Loan Repayment Type from Mortgage Deed final offers.
2025 Week 31
- Improved AI search filter for more relevant results.
2025 Week 30
- Address details - Street, House Number, House Number Suffix, Zip Code, City - are now included for all address extractions across all document types. This data can be found under 'AdressDetails' property in API response.
2025 Week 29
- Improved monthly income extraction for Pension Benefit Decision documents.
2025 Week 28
- Improved the startup speed of document archive processing through AI Integrations.
- Enabled insurer company and policy number data extractions for all Policy Document types.
2025 Week 27
- Improved monthly income extraction from Monthly Payment Specification documents, particularly for AOW specifications.
- AI classification now recognizes new document type: "Financing plan".
2025 Week 26
- Improved start date extraction from the latest residence permit format.
2025 Week 23
- AI API clients can now customize data retention settings on their own.
2025 Week 22
- Added support for Salary Annual Statement as a recognizable document type, extracting the following 6 data points: Employer Name, Employee Name, Employee Identifier, Yearly Income, Withheld Income Tax, Tax Credit.
- Added support for Student Loan as a recognizable document type, extracting the following 9 data points: Full Name, Personal Address, Date Of Document, Student Loan Balance, Loan Repayment Type, Interest Rate, Interest Rate End Date, Loan Remaining Term, Loan Repayment Amount.
- Added support for Bank Annual Statement as a recognizable document type, extracting the following 5 data points: Full Name, Bank Name, Initial Amount, Final Amount, Statement Year.
- Added support for Bank Statement as a recognizable document type, extracting the following 5 data points: Full Name, Bank Name, Personal IBAN, Account Final Amount, Date Of Document. This document type is also sub-categorized into: Checking account statement, Savings account statement, Investment account statement.
2025 Week 18
- Updated data extractions for the "BKR outcome" document type.
2025 Week 14
- Expanded the cut-out area of passports to meet new legal requirements.
- Extended ID document cut-out to support newest Dutch passport's back page.
2025 Week 12
- Added the ability to define a custom render DPI for rendering PDF page into an image.
- Improved signature detection to recognize a broader range of signatures placed in different locations within the document.
2025 Week 10
- Added the ability to convert HTML document to PDF.
- Removed "Renovation specification" (Dutch: Verbouwingsspecificatie) document type.
- All documents previously identified as "Renovation specification" are now classified as "Additional work specification" (Dutch: Meerwerkspecificatie).
2025 Week 7
- Added the ability to extract data from the "Mixed" document type when it contains other known document types.
2025 Week 4
- Improved BSN blurring for incorrectly formatted BSNs near the keyword 'BSN'.
2024 Week 50
- Extended date parser to support three date formats: 31122005, 31 12 2005, and Dec 31, 2005.
2024 Week 48
- Added a new data point for Pension Name, extracted from classified documents such as Pension Overview (Dutch: MijnPensioenoverzicht).
2024 Week 46
- Added Full Name extraction from non-MRZ zone of Dutch ID documents, useful for cases where a person's name does not fit in MRZ.
- Duplicate extraction filter has been updated to correctly handle cases where the original text layer was in the wrong location.
- Fixed an issue with PDF layout calculations to better handle documents with unusual or extended page sizes.
2024 Week 44
- Introduced an option to extract the full MRZ zone without parsing, displaying the data exactly as it appears in ID document.
- Added an option to exclude attachments when converting emails to PDFs.
- Improved Signature and Checkbox classification models, resulting in better parsing of Declaration of Work documents (affecting multiple data points) and more accurate signature detection on various documents, such as Declaration of Work, Mortgage Deed, Final Offer, and more.
2024 Week 42
- Added a new data point for Place of Birth, extracted from the non-MRZ zone of Dutch ID documents.
- Added new data points for Employee Identifier, End Date of Employment, Weekly Hours, Employment Type, Employee IBAN, extracted from the classified documents such as Salary Slip (Dutch: Salarisstrook).
2024 Week 40
- Increased file merge limit from 20 to 150 files.
- Added ability to convert Excel documents to PDF.
- Added new data points for Surname, Initials, Date of Birth, Date of Document, Year of Tax Return, Incoming from Company Profits, Lease Hold Deduction, extracted from classified documents such as Income Tax Report (Dutch: Aangifte).
2024 Week 38
- Added new data points for Collateral Address, Purchase Price, Passing Date, Date of Bank Guarantee, Date of Resolutive Conditions, buyer's and seller's information (Surname, Initials, Date of Birth, Address, Telephone Number, E-mail, Marital Status), extracted from classified documents such as Purchase Agreement (Dutch: Koopovereenkomst).
2024 Week 36
- Deed of Sale (Dutch: Koopakte) document type has been replaced with three more accurate document types: Deed of Transfer (Dutch: Leveringsakte), Purchase Contract (Dutch: Koopovereenkomst), and Construction Agreement (Dutch: Aanneemovereenkomst).
- Added new data points for Part Time Percentage and Pension Contribution Amount, extracted from classified documents such as Salary Slip (Dutch: Salarisstrook).
2024 Week 34
- Added a new data point for cross-checking. Yearly Income can now be cross-matched between Declaration of Work and Salary Slips.
- Added new data points for Employment Continuation Signature, Employment Continuation Name of Signatory, Employment Continuation Type, Employment Continuation Amendment State, Model Date, Employer Address, Job Position, KVK Number, Trial Period Exists, Trial Period Passed, Planned Reorganization, extracted from classified documents such as Declaration of work (Dutch: Werkgeversverklaring).
- Improved detection of small or unclear faces in ID documents.
2024 Week 32
- Added new data points extractions from document types:
- Surname, Initials, Date of Birth - from UWV Insurance (Dutch: UWV verzekeringsbericht).
- Surname, Initials, Date of Birth - from Pension Overview (Dutch: MijnPensioenoverzicht).
2024 Week 28
- Added a new data point for cross-checking. Date of Birth, extracted from both Declaration of Work and ID documents, can now be compared.
- Added a Checkbox classifier that locates and classifies checkboxes into two categories: "Checked" and "Not Checked." This internal model aims to improve the accuracy of Declaration of Work parsing.
2024 Week 26
- Added ability to identify which specific documents caused the failure when merging them into a single PDF.
- Added PII (personal identification information) extraction from Belgium ID.
- Enabled optional processing of XFA PDFs.
2024 Week 22
- Added the ability to cross-match data points within mortgage documents. Employee Name and Employee Address can be cross-checked between Salary Slip and Declaration of Work.
- Improved text layer application on PDFs by resolving occasional issues where spaces between words or characters were omitted.
2024 Week 20
- Added new data points extractions from document types:
- Employee Name, Employee Address, Date of Birth, Date of Employment, Shareholder Type and Name of Signature - from Declaration of work (Dutch: Werkgeversverklaring).
- Employee Name, Employee Address, Date of Birth, Date of Employment - from Salary slip (Dutch: Salarisstrook).
- Improved data extraction accuracy on ID documents and passports by filtering out lines with fewer than 15 characters.
2024 Week 18
- Enabled processing of document groups, streamlining batch operations.
- Added an option to process signed PDFs.
- Added new data points extractions from document types:
- Incorporation Date and Last Amendment Date - from Excerpt of Chamber of Commerce (Dutch: Uittreksel KVK).
- Buyer's Address - from Deed of Sale (Dutch: Koopakte).
- Lender, Loan Part Number, Current Debt Remaining, Loan Amount, Loan Repayment Type - from Mortgage debt statement (Dutch: Schuldrestopgave hypotheek).
2024 Week 16
- Added the ability to configure (per workflow) which document types and/or data points are used.
- Small photos in passports are now detected and blurred.
2024 Week 14
- Added the ability to name workflows, improving maintainability when managing a larger number of them.
- Added a new data point for Issuing Authority, extracted from Dutch ID documents.
- Enhanced BSN detection on damaged MRZ, ensuring it is now blurred.
2024 Week 12
- Improved the BSN detector to enhance its accuracy and robustness.
2024 Week 10
- Added new data points extractions from document types:
- Date of Document - from Articles of Association (Dutch: Statuten).
- Start Date - from Dutch Residence Permit.
- AI classification now recognizes new document type: Mortgage Debt Statement (Dutch: Schuldrestopgave hypotheek).
2024 Week 8
- Developed a duplicate extraction filter to eliminate redundant data points when the same information is extracted multiple times.
- Created a mechanism to extract tables and utilize their data to deliver essential information to users.
2024 Week 4
- Added new data points extractions from document types:
- Monthly Income and Yearly Income - from Pension Benefit Decision (Dutch: UWV toekenningbesluit uitkering).
- Income Type and Yearly Income - from Declaration of Work (Dutch: Werkgeversverklaring).
- Monthly Income, extracted - from Monthly Payment Specification (Dutch: UWV maandspecificatie uitkering).
2024 Week 2
- Introduced an option to specify accepted file formats at the workflow level, helping users prevent the processing of unintended or unnecessary documents.
- Improved filtering of BSN false positives to reduce incorrect blurring of areas in ID documents.