TenderSRB
Published:
AI now reads all tender documents
AI analysis now reads documents in Word 97–2003 (.doc), OpenDocument (.odt), RTF and ZIP archives, in addition to the PDF, DOCX and XLSX files it already read. A check on 19.10. showed that on the jnportal portal, the technical specification for 18% of tenders was only in such files; the model now sees it.
Duplicate files in lots no longer push out unique documents. Scanned documents without a text layer are marked; text recognition (OCR) is not performed.