IDP Case Study: Agribologna Automates 130,000 Agri-Food Documents a Year with AI

130KDocuments/year

-60%Processing time

99.8%Accuracy

The challenge

Agribologna, a leading fruit and vegetable cooperative in Emilia-Romagna, handles over 130,000 documents per year across its sales and procurement cycles: inbound and outbound DDT shipping documents, customer orders, price and quantity adjustments, equipment rental invoices, and logistics service invoices. Manual entry into their ERP required a dedicated team and caused recurring errors on item codes, quantities, and order references.

The solution

Data Alchemy automatically captures all documents — from email, scanners, and supplier portals — and extracts structured data with AI. The system identifies the document type, maps fields to the ERP, and runs automatic validation on every line by cross-referencing item master data, price lists, and existing orders. Adjustments are reconciled against the original DDTs, while rental and service invoices are matched to active contracts through automatic invoice reconciliation. Where a purchase order and a delivery note exist, the platform performs automatic three-way matching (invoice / purchase order / DDT), flagging only the documents that fail a check for human review — everything else is posted to the ERP in about 3 seconds.

Documents processed

▶DDT — inbound and outbound goods▶Customer orders — via email, PDF, portal▶Adjustments — price and quantity changes▶Rental invoices — vehicles and equipment▶Service invoices — logistics and transport

“Data Alchemy has transformed the way we manage documents. What used to take hours of manual data entry now happens automatically, with an accuracy that surpasses human performance. Our team can finally focus on value-added activities.”

— CIO, Agribologna

How Agribologna Went Live in 5 Working Days

Unlike traditional OCR projects that require weeks of template design, the Agribologna deployment required no document templates at all. The rollout followed four steps:

1. 01

   Day 1 — Document capture setup

   A dedicated mailbox on Microsoft 365 was connected to Data Alchemy, so inbound DDTs, customer orders and supplier invoices arriving by email are captured automatically, with AI spam filtering discarding non-relevant attachments. Scanner and supplier-portal sources were added to the same intake queue.

2. 02

   Days 2–3 — ERP integration

   Data Alchemy was connected to Agribologna's ERP through a native two-way connector, exposing item master data, price lists and open orders to the extraction engine. No middleware and no manual export/import were required.

3. 03

   Day 4 — Validation rules

   Every extracted line is checked in real time against the ERP: item codes must exist in the master data, prices must match the active price list, and quantities are reconciled against the referenced order or original DDT. Documents that fail a check are routed to a human review queue instead of being posted with errors.

4. 04

   Day 5 — Operator training and go-live

   The administration and logistics teams were trained on the review queue. From day one, documents that pass validation are posted to the ERP without human intervention, in roughly 3 seconds each.

Because the model is pay-per-use (from €0.50 down to €0.35 per document, with no monthly subscription), Agribologna paid nothing for unused capacity during the ramp-up phase.