Data Alchemy — Software IDP con AI
Gruppo Montenegro
Case study

10,000 orders a year, 150 different formats: how Gruppo Montenegro automated order entry with Data Alchemy

From the PDF in the mailbox to the order in Salesforce, without retyping a single field. Even at night and at weekends.

Key figures · Food & beverages · Salesforce

Orders a year
~10,000
Different formats
150+
Average time per order: −71% (Montenegro's estimates)
7 to 2 min

Case sheet

Status:In production

Customer
Gruppo Montenegro
Sector
Food and alcoholic beverages
Based in
Zola Predosa (Bologna), Italy
Process automated
Customer order entry
Documents
Customer orders as PDFs, in over 150 different formats
Capture channel
Email
Volumes
Around 10,000 orders a year
Key results
  • Average time per order from 7 to 2 minutes (−71%, according to Montenegro's estimates)
  • Systematic check of the data against Salesforce before the order is generated
  • Orders processed 24 hours a day, 365 days a year

Who Gruppo Montenegro is

Gruppo Montenegro was founded in Bologna in 1885, when Stanislao Cobianchi started his business as a liqueur maker. Today it is a leading Italian company in the food and alcoholic beverages sectors, headquartered in Zola Predosa (BO). Its portfolio includes brands that are part of Italian culture: Amaro Montenegro, Vecchia Romagna, Rosso Antico, Bonomelli, Cannamela, Cuore and Polenta Valsugana, as well as Select Aperitivo and Pampero rum, acquired from Diageo in 2024. With five production plants in Italy, the Group distributes its products in more than 70 countries.

A company with over 140 years of history that has chosen to invest in Artificial Intelligence applied to business processes.

The problem

Every company that sells through large-scale retail, the Ho.Re.Ca. channel or major accounts knows the situation well: orders arrive by email, as PDFs, and every customer uses its own format. Someone has to open them, read them and copy them by hand into the management system.

At Gruppo Montenegro this work fell to the Key Account Managers. The numbers give a sense of the problem: around 10,000 orders a year, in over 150 different formats, with an average entry time of 7 minutes per order. Time taken away from the customer relationship and, as with any data entry activity, exposed to errors that turn into returns and credit notes.

Montenegro's Digital Innovation Team chose Data Alchemy, our Intelligent Document Processing platform, to eliminate this manual step.

How the flow works

The strength of the project is not data extraction from the PDF alone, but the native integration with Salesforce, which turns a document into a validated order assigned to the right person:

  1. 01

    Receipt. Data Alchemy directly receives the emails containing customer orders.

  2. 02

    Customer recognition. Using Large Language Models and Computer Vision, the platform reads the attached document and identifies the customer, whatever the format.

  3. 03

    Check against Salesforce. Thanks to the native connector, the extracted data is compared with the data held in Salesforce: customer code, price list, item codes, prices and sales manager.

  4. 04

    Assignment. The processed document is associated with the relevant Key Account Manager, who is the only person who can see it within Data Alchemy.

  5. 05

    Exception handling. If information is missing from the order, the system sends a completion notification to the person responsible, who steps in only where needed.

  6. 06

    Order creation. Once complete, the order is sent to Salesforce automatically.

The KAM thus moves from data entry operator to supervisor: they check, complete where necessary, and the rest happens automatically.

A colleague who works around the clock, 365 days a year

Orders do not only arrive during office hours. Many customers send them in the evening, at weekends or in the early hours of the morning, and until recently those documents piled up in the mailbox waiting for someone to process them.

Data Alchemy keeps no hours: it processes orders the moment they arrive, 24 hours a day, 365 days a year. When the Key Account Managers arrive at work at the agreed time, they find the complete orders already loaded into Salesforce and, in their own reserved area, only those that need completing. No backlog to clear on Monday morning, no day that starts late.

The result is a steadier flow along the whole chain: orders enter the system sooner, logistics can plan sooner, and people start the day by focusing straight away on value-adding work.

The operational benefits

According to Montenegro's estimates, the average time per order drops from 7 to 2 minutes, a 71% reduction.

The change is not only about speed. The very nature of the activity changes: the Key Account Manager no longer transcribes data, but checks and completes it when needed. Attention shifts from repetitive operations to the customer relationship, which is the job that role exists for.

Data quality

Manual transcription, however careful, inevitably involves errors: a wrong item code, a price not aligned with the price list, a transposed quantity. Every error spreads downstream, generating returns, credit notes and logistics rework.

With Data Alchemy, the extracted data is automatically compared with the data held in Salesforce before the order is generated. The check no longer depends on the attention of whoever is typing at a peak moment: it is systematic and applied to every document in the same way. Where the data is incomplete, the system stops and flags it, instead of guessing.

A solution that scales

The process is independent of document format and volume: the original 150 formats do not require 150 configurations, and a seasonal peak in orders does not turn into a peak in manual work or processing queues. The same extraction, checking and routing logic can be applied to other document types that share the same pattern (order confirmations, delivery notes, supplier invoices), reusing the integration already in production.

What this project teaches us

Generative AI creates real value when it does not remain an isolated experiment, but plugs into the systems the company already uses. Reading a PDF is relatively simple today; what makes the difference is checking the data against master data, price lists and item codes, routing the document to the right person and closing the loop all the way to the order in Salesforce, seamlessly.

Our thanks to the Digital Innovation Team at Gruppo Montenegro for their trust and collaboration at every stage of the project.

If orders at your company (or invoices, delivery notes, order confirmations) still go through manual copy and paste, let's talk.