Document Intelligence
This example processes synthetic invoices and a purchase order through explicit ingestion, classification, extraction, reconciliation, and reporting activities. It verifies totals and duplicate invoice identifiers rather than stopping after text extraction.
ingestclassifyextractreconcilereport
Inputs
- Versioned semi-structured invoices
- Purchase-order fixture
Outputs
- Document inventory
- Classifications
- Extracted fields and line items
- Reconciliation JSON and CSV
- Aggregate report
Provenance to preserve
- Source hashes
- Document-to-record mapping
- Balance checks
- Duplicate checks
- Gross total
Run locally
git clone https://github.com/UFFeScience/akoflow-examples.git
cd akoflow-examples/document-ai/document-intelligence
./run.sh
The fixture is intentionally small and private-data-free. Replace data/documents with a larger local corpus while preserving the same artifact contract.