From emails to orders in just a few minutes: New Penta's AI case study
How can you transform emails, PDFs, scans, and unstructured requests into orders ready for verification? In this case study, we detail New Penta’s digital transformation journey and the project developed with Zero11: an AI solution integrated into business processes that can automatically interpret order requests, recognize customers, products, and quantities, and support customer service in generating pre-orders. A concrete example of AI applied to operations, with measurable results: 70–80% of emails are correctly interpreted, and order processing times have been cut by more than half. Translated with DeepL.com (free version)
18 December 2025
What you'll learn
- How to Turn Emails and Unstructured Documents into Orders Using AI
- How to Integrate Large Language Models, Computer Vision, and RAG into Business Processes
- Why Use a Human-in-the-Loop Approach
- How to protect sensitive data with a privacy-first AI architecture
- How to integrate AI with CRM, e-commerce, and order management
- How New Penta has more than halved order processing times
- How to identify business processes suitable for intelligent automation
- How to evolve from a single AI use case toward a transformation roadmap
About
When AI Truly Becomes Part of Business Processes
Artificial intelligence generates value when it moves from experimentation to production and is applied to real-world problems.
This is what happened at New Penta, a company within the Named Group, as part of a broader digital transformation initiative developed in collaboration with Zero11.
The Problem
A significant portion of orders continued to arrive through traditional channels: emails, phone calls, PDF documents, scans, photographs, and files in a wide variety of formats.
As a result, the customer service team had to:
- manually read and interpret requests;
- identify the customer or patient;
- recognize products and quantities;
- verify the accuracy of the information;
- enter the order into the company’s systems.
This process was critical to the business but was characterized by repetitive tasks, long processing times, and unstructured information.
The Solution
Zero11 developed an AI solution capable of reading emails and attachments, interpreting them, and converting them into potential orders within the company’s platform.
The solution combines:
- Large Language Model;
- computer vision capabilities to understand documents and images;
- open-weight models;
- RAG for product recognition;
- integration with HubSpot;
- integration with order management systems;
- confidence scores to support operator verification;
- human-in-the-loop;
- privacy-first architecture.
The goal is not to eliminate human oversight, but to allow people to focus on exceptions and higher-value tasks.
The Results
The system is now able to correctly interpret approximately 70–80% of emails and has made it possible to more than halve order processing times.
At the same time, the digitization initiative has helped increase the share of orders coming from e-commerce from approximately 20% to 50% in two years.
Beyond Order Automation
The project also serves as a platform on which to build new developments:
- chatbots and AI agents for support;
- new touchpoints such as WhatsApp;
- redesign of the e-commerce experience;
- analysis of anonymized scientific data;
- digitization of signature processes;
- new digital services for customers and patients.
A concrete example of how AI can become a structural component of business transformation.
Watch the webinar and discover how to transform a manual process into a measurable, manageable AI use case that is truly integrated into the business.
Translated with DeepL.com (free version)
Speakers
- AB
Andrea Bosio
AI Evangelist · Aiability
- EF
Edoardo Franco
AI Engineering · Aiability
Related content
The Future of B2B Commerce: Strategies, Technology, and AI with Jason Greenwood
B2B commerce cannot be treated as B2C e-commerce with a few additional features. The buyer, processes, payments, pricing, logistics, credit, marketing, and shopping experience are all different. In this international interview for B2B Commerce Corner, Andrea Bosio discusses the origins and vision of Rewix and shares his thoughts on the future of digital commerce between businesses.
59 min
Watch →
Andrea Bosio spricht über Rewix und erläutert, warum Prozesse, UX, Zahlungsverkehr, Preisgestaltung, CRM und Technologie im B2B-Handel einen spezifischen Ansatz erfordern.
Artificial intelligence is profoundly changing the way companies and marketers analyze data, understand people, and design increasingly personalized customer experiences. In this webinar, we’ll explore the intersection of AI and MarTech, from the evolution of traditional inbound and outbound marketing models to the use of analytics, natural language processing, knowledge graphs, recommendation systems, and AI agents. A journey to understand how data, technology, and artificial intelligence can work together to make marketing smarter, more contextual, and more focused on the customer’s real needs.
58 min
Watch →
AI Act 2026: What Really Changes for Companies, Providers, and Deployers
On August 2, 2026, the AI Act enters a new phase of implementation. But what does this really mean for those who develop, integrate, or use artificial intelligence systems? In this webinar, Andrea Bosio of Zero11 speaks with Maria Chiara of Studio Torta to clarify the responsibilities, obligations, and risks introduced by the European regulation. Providers, deployers, data quality, high-risk systems, AI literacy, generative models, governance, documentation, and penalties: a practical discussion to understand what companies need to do today to prepare.
47 min
Watch →