We’re developing Pacioli, a Logic Toolkit for XBRL-Based Digital Financial Reporting. It is a new software infrastructure for AI based report checking, based on a robust logic programming engine, providing:
- Loading of reports and their Discoverable Taxonomy Sets into Prolog-based compact logical representations
- Formula processor, with auditing and explanation capabilities
- Combination of a report with local user alterations of its formulas and facts
- SBRM predicates
- Overall “rules as code” approach, for easier evolution and maintenance
You can try the current version on existing (web published) XBRL reports:
- Interactive notebooks in our demo server, including taxonomy viewers, report alterations editor, and further technical information.
- The simple form below:
To process a report now, please fill your email (for diagnostic purposes only) and paste the URL of a XBRL report instance (this must be within the xbrlsite.azurewebsites.net or www.xbrlsite.com domains for the time being):
You should see the result here in a few seconds, pending network traffic. There’s more functionality at the notebooks site.
A core subset of XBRL and extensions is supported at this time, the sufficient to handle the PROOF report example and most public filings. Subsequent development will include:
- Add more model patterns e.g. for SBRM
- Definition of higher level rules for report checking, in PROLOG and Logical English
- Enhanced explanations/auditing
- Integration into specific user applications
- Reporting quality control
For more information please see the referred notebook or contact licensing@logicalcontracts.com