Automatically match extracted data to your system of record

AI fuzzy matching compares extracted document data against your system of record using contextual understanding rather than exact string matching. When a purchase order shows “Widget Assembly Unit” and your ERP lists “Widget Assy Unit,” fuzzy matching recognizes these as the same item based on semantic similarity, abbreviation patterns, and surrounding context. This resolves the data inconsistencies that make traditional VLOOKUP-based reconciliation fail on real-world documents.

Yes. You can configure the confidence threshold that determines when a match is automatically accepted versus flagged for human review. A higher threshold means more records go to manual review but fewer errors pass through. A lower threshold means more automatic matching but higher risk of false positives. Most teams start with a conservative threshold and adjust as they build confidence in the system’s accuracy for their specific document types.

When a match cannot be found, the record is flagged as an exception and routed to your team for manual review. The system provides the extracted data alongside potential partial matches and the reason the automatic match failed—whether it was a missing reference number, a data format mismatch, or no corresponding record in your system. This gives your team the context they need to resolve the exception quickly rather than starting from scratch.

Automated data reconciliation works with both standard ERP systems and proprietary databases. Lido exports reconciled data in structured formats including CSV, Excel, and JSON that can be imported into any system. For direct integration, Lido connects to NetSuite, Salesforce, QuickBooks, and other platforms via API. If you use a proprietary or industry-specific system, the structured export ensures your reconciliation data flows in without manual reformatting.