Capture the information you need in any type of documents


  • reciTAL EXTRACT has an understanding of all types documents, long and short, as a human would :
  • It locates and reads entities (names, addresses, amounts), text, titles and their hierarchy, images, tables and forms.
  • Basic business rules are then implemented in a few minutes to configure your data points.
  • Depending on the level of confidence reached, the extracted data are validated or not by an operator.


  • Bulk extraction data from document flows for Back, Front and Middle Offices
  • Software editors integrating IDP functionalities
  • Integrators, consultants with needs to build complete workflows integrating IDP functionalities

R.O.I ?

reciTAL EXTRACT is the fundamental IDP module for the automation of your document flows :

  • Extraction without training is immediately operational
  • The configuration of a new type of document is done with unparalleled simplicity and speed
  • Any type of documents handled, including the most complex
  • Validated data is used for continuous performance improvement

Understanding of all types of documents

  • OCR
  • Multilingual
  • Structured documents (ID card, passports, driving license, …),
    semi-structured (invoices, purchase orders, tax assessment, …), unstructured (reports, contracts, …)
  • Identification by default of all the characteristics and data of the document :
    • Structure (title, paragraph, hierarchy, image, footer ..)
    • Entities (amounts, names, addresses, dates, organizations …)
    • Tables (reading table details, cell by cell)
    • Forms (keys / values)

Automate the capture of data points from your documents

  • Extraction of data points with confidence thresholds :
    • automated data transfer to your ERP, databases, CRM and other systems
    • if necessary, post processing error correction interface with prefilled fields (data entry time reduced by 10)
  • Monitoring of activity : documents received, being processed, processed

How does it work ?


  • If your document is one we already know and manage, you have nothing to do and can directly start extracting your data!
  • If you have a new type of document, you have to provide three examples to reciTAL or to your integrator, and the configuration is done in a few hours maximum.


  • If the confidence threshold is high, the process is completely automated. If necessary, use of the post processing error correction interface with prefilled fields
  • Integration of EXTRACT into your system directly through API connector or through your workflow module.

Hosting : use EXTRACT wherever you want

reciTAL mutualized Cloud

SaaS offer. Secure, redundant environment, hosted in France (OVH / Equinix), managed by the reciTAL team.

Your Cloud

OVH, AWS, Azure… the reciTAL team guides you in your implementation.

Your infra

The reciTAL team guides you through the deployment in your internal Kubernetes or via Docker Compose.