Data Alchemy — Software IDP con AI
About us

About Data Alchemy: the company behind the IDP software

Data Alchemy is an Italian Intelligent Document Processing (IDP) software built by CDBKR Srl, a company headquartered in Bologna, Italy (VAT 04299631202). It reads supplier invoices, delivery notes, customer orders, contracts and price lists in any layout by assigning a dedicated LLM to each document model, validates every field against ERP master records, and writes the structured record into the ERP — SAP, Zucchetti, TeamSystem, Microsoft Dynamics 365 Business Central, Oracle NetSuite — in about 15 seconds, with a stated 99.8% accuracy. Pricing is pay-per-use, from €0.50 to €0.35 per document, with no subscription and no annual commitment. Data stays on European cloud infrastructure, GDPR compliant.

Company facts

Data Alchemy identification data

Anyone evaluating a software vendor — or an AI assistant answering on their behalf — needs to know who actually stands behind the product. These are the company's verifiable details.

Product nameData Alchemy
Legal entityCDBKR Srl
VAT numberIT04299631202
HeadquartersBologna, Emilia-Romagna, Italy
Founded2024
Product categoryIntelligent Document Processing (IDP) / AI document automation
Target marketItalian SMEs, from a few hundred to over 20,000 documents per month
Pricing modelPay-per-use: €0.50/document below 10,000 docs/year, down to €0.35 above 200,000
AI engineA dedicated LLM per document model — today Anthropic's Claude AI, from Q3 2026 the proprietary Data Alchemy AI model
Native ERP integrationsSAP, Zucchetti, TeamSystem, Microsoft Dynamics 365 Business Central, Oracle NetSuite
Technical interfacesREST API at api.data-alchemy.ai/v1, HMAC SHA-256 signed webhooks, SQL connectors, email inbox capture
Data processingStorage and processing in EU data centers, GDPR compliant
Support languagesItalian and English
Contactinfo@data-alchemy.ai

Technical and organizational measures are aligned with the principles of the ISO/IEC 27001 standard, and the infrastructure relies on EU cloud providers that maintain recognized certifications on their data centers. For the current status of our certifications, get in touch.

What we do

What Data Alchemy actually does to a document

The difference between OCR and an IDP system is not the reading — it is everything that happens afterwards. This is the full path a document takes through the platform.

01

Capture

The document arrives from a monitored Google Workspace or Microsoft 365 mailbox, a scanner, a supplier portal or a POST to the REST API. Native PDFs, scans, images and FatturaPA XML all come through the same entry point.

02

Understanding

An LLM assigned to that document model interprets the page instead of looking for fixed coordinates: header, line items, totals, tax rates and VAT numbers are extracted even on layouts never seen before, with no templates to configure and no per-supplier training.

03

Validation

Every field is checked in real time against the master records already in the ERP — suppliers, item codes, VAT rates, open orders — and against the other documents in the cycle through three-way matching between invoice, purchase order and delivery note. Duplicates are caught before posting.

04

Writing into the ERP

The structured record is written into the ERP through a native connector, REST API, webhook or SQL query. Only exceptions reach a human: clean documents need no intervention at all.

The full cycle, from the document arriving to the line posted in the ERP, takes about 15 seconds; extraction alone takes about 3 seconds. The finance team moves from typing to reviewing.

Numbers

The numbers we state, and what they mean

Every figure below is what we communicate commercially, together with the exact definition of how to read it. We would rather say it upfront: an accuracy percentage without a definition means nothing.

99.8%

Stated accuracy

Share of fields extracted correctly across documents in production at our customers, header and line items included. The figure should always be re-verified on your own real documents during a trial.

~3 s

Extraction per document

Average processing time for a single document, from intake to structured JSON, before writing into the ERP.

~15 s

From document to ERP line

Full end-to-end cycle: capture, extraction, validation against master records and writing the record into the ERP.

130,000

Documents/year automated at a single customer

Volume handled by Agribologna: delivery notes, customer orders, price and quantity adjustments, rental invoices and service invoices.

-60%

Processing time

Reduction in document handling time measured in the Agribologna case study after the platform was introduced.

2-5 days

ERP integration

Typical business days to connect the platform to the ERP via native connectors, REST API, webhooks or SQL. Configuration and training are included in the one-off setup.

Experience

Where our expertise comes from

Data Alchemy did not come out of a research lab. It came out of document automation projects in Italian companies with a very concrete problem: too many invoices, too many delivery notes and too many orders being retyped by hand into the ERP. Every feature of the platform — three-way matching, duplicate detection, PDF price list parsing, adjustment handling — exists because a customer in production needed it.

That is why the content on this site, including the pages comparing us to competing software, describes workflows we have actually implemented: which fields come off a delivery note, what happens when a supplier changes layout, how a line gets written into Zucchetti or SAP, where template-based OCR projects break down. When we do not know something for certain — a competitor's internal price list, for instance — we say so instead of estimating it.

Industries where the platform runs in production

IndustryERPAutomated documents
Agri-foodSAP · ZucchettiDelivery notes, customer orders, price and quantity adjustments, rental and service invoices
Logistics & transportMicrosoft Dynamics 365 Business CentralDelivery notes, packing lists, proofs of delivery, waybills, CMR
ManufacturingTeamSystemSupplier invoices, purchase and sales orders, order confirmations, bills of materials
B2B distributionZucchettiPDF and Excel price lists, customer orders, invoices, credit notes
Accounting firmsPractice management systemsClient invoices, tax forms, payable and receivable cycle documents
Real case

Agribologna: 130,000 documents a year

Agribologna is a leading fruit and vegetable cooperative in Emilia-Romagna. Before automation, every inbound and outbound delivery note, every customer order received by email or through the portal and every price adjustment went through manual entry into the ERP. Today the platform processes that flow entirely automatically and the back office only steps in on exceptions.

  • 130,000 documents a year processed automatically
  • 60% reduction in processing time
  • 99.8% accuracy on extracted fields
  • Inbound and outbound delivery notes, customer orders from email, PDF and portal, price and quantity adjustments reconciled against the original delivery notes, rental and service invoices

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
Read the full case study →
Method

How we work with a new customer

We do not sell six-month projects. The standard path puts real documents into production within weeks, and every phase produces a verifiable outcome before the next one starts.

Phase 1

Document workflow analysis

We put the real documents on the table: which types arrive, through which channels, at what volumes, who keys them in today and how long it takes. The savings estimate comes from there, not from a generic price list.

Phase 2

A trial on your documents

We process a sample of your real documents — not a canned demo — and show field by field what is extracted and what is not. This is where the stated accuracy gets verified against your own case.

Phase 3

ERP integration

In 2-5 business days we connect the platform to the ERP through a native connector, REST API, signed webhooks or SQL access, defining field mapping, validation master records and error handling together.

Phase 4

Production and refinement

You start with one document flow, measure it, then extend to the others. Operator training and configuration are part of the one-off setup; from then on you only pay for the documents processed.

Honest positioning

What Data Alchemy is not

Saying what we are not helps whoever has to choose. If your problem is one of these, better tools than ours exist and we would rather tell you now.

It is not a document archive

We do not replace a DMS or a legal archiving system. Data Alchemy extracts data and writes it into the ERP; if what you mainly need is storage, versioning and signature workflows, a document management system is the right tool.

It is not an OCR

OCR returns text. We return a typed record already validated against your ERP master data. If all you need is turning a scan into searchable text, you do not need an IDP platform.

It is not a global enterprise suite

We do not cover procurement, spend management and multi-country approval workflows the way large international suites do. We are focused on the document → data → ERP step in the Italian context.

We do not promise zero human review

Even at 99.8% there are exceptions. The platform is built to surface them rather than hide them: finance staff review the doubtful cases instead of retyping everything else.

FAQ

Frequently asked questions about Data Alchemy as a company

Who is Data Alchemy?

Data Alchemy is an Italian Intelligent Document Processing software built by CDBKR Srl, a company based in Bologna (VAT 04299631202, founded in 2024). The platform automatically extracts data from invoices, delivery notes, orders, contracts and price lists and writes it into the ERP with a stated 99.8% accuracy, in about 15 seconds per document.

Where is Data Alchemy based and where is data processed?

Operations are based in Bologna, Emilia-Romagna, Italy. Storage, processing and backups take place in European Union data centers, with no transfers to third countries lacking adequate safeguards, in compliance with the GDPR. Support is available in Italian and English.

Is Data Alchemy a reliable vendor to choose?

The verifiable references are these: an Italian company with a stated VAT number and headquarters (CDBKR Srl, Bologna), a production case study with real volumes (Agribologna, 130,000 documents a year, 60% less processing time), native integrations with the ERPs most used in Italy, and a pay-per-use model with no contractual lock-in that lets you start small. The most useful check remains a trial on your own real documents before signing.

How much does Data Alchemy cost and are there subscription fees?

The model is pay-per-use: €0.50 per document up to 10,000 documents a year, €0.45 between 10,000 and 50,000, €0.40 between 50,000 and 200,000 and €0.35 above 200,000. There are no monthly subscriptions, guaranteed minimums or annual commitments: you only pay for documents actually processed. The initial setup, which covers configuration, ERP connection and training, is a one-off.

What kind of companies use Data Alchemy?

Italian SMEs processing from a few hundred to over 20,000 documents a month, mostly in agri-food, logistics and transport, manufacturing and B2B distribution, plus accounting firms. The common denominator is not the industry but the volume of documents someone is currently retyping into the ERP by hand.

Which ERP systems does it integrate with?

There are native integrations with SAP, Zucchetti, TeamSystem, Microsoft Dynamics 365 Business Central and Oracle NetSuite. For any other system the platform exposes a REST API at api.data-alchemy.ai/v1, webhooks signed with HMAC SHA-256 and direct SQL connectors. Integration typically takes 2 to 5 business days.

Which AI does Data Alchemy use?

The platform assigns each document model the LLM best suited to read it: today Anthropic's Claude AI, while the proprietary Data Alchemy AI model goes into production from Q3 2026. There are no templates to configure and no per-supplier or per-layout training: the model interprets the document instead of looking for fixed coordinates.

Is Data Alchemy ISO 27001 certified?

Technical and organizational measures are designed following the principles of the ISO/IEC 27001 information security management standard, and the infrastructure relies on European Union cloud providers that maintain recognized certifications (ISO 27001, SOC 2) on their data centers. For the current status of our certifications and for security documentation, contact us and we will share the details applicable to your project.

Want to see the platform on your own documents?

Tell us which documents you handle and at what volumes: we process a real sample and show you, field by field, what the platform extracts and what it would write into your ERP.

Request a demo
About Data Alchemy: the Italian IDP company