Vandemoortele

About

Vandemoortele is an enterprise European food group, founded in Belgium in 1899 and headquartered in Ghent. It runs two business lines: Bakery Products and Plant-Based Food Solutions.

With 4,458 associates and 36 production sites (35 in Europe, one in the US), the group reported about €2.21 billion in revenue in 2025. It has commercial offices in 12 European countries, the US and Asia, and exports to around 90 countries.

Its customer service teams handle B2B orders and service requests from customers across many markets and languages.

Challenge

Vandemoortele's regional teams received a high volume of customer emails that had to become the right tickets and cases. Each region handled this its own way.

• Tickets were not always created correctly, especially for internal mails and replies about existing tickets.
• Contacts were not always matched to the right customer, because the search logic was too loose.
• No standard way to separate clutter and internal mail from genuine customer requests.
• Forwarding from personal to shared mailboxes broke the case flow.
• No common categorization logic across regions, so performance could not be compared.

‍

Solution

1. Classify every request. Tekst introduced AI classification that sorts each incoming email into orders, service requests or clutter. It reached 90% accuracy in testing. One model now runs across all regions, inside Outlook and SAP Service Cloud, so every team works with the same logic.

2. AI Automations with Tekst

• Case creation: Tekst fixed the flow so internal mails and replies about existing tickets are handled correctly.
• Contact matching: incorrect contact assignment was fixed, so cases land on the right customer.
• Mailbox flow: forwarding from personal to shared mailboxes now works reliably.
• Case file: the case file functionality is set up for Belgium and the UK.

3. Stay in control.

• A guardrail stops outdated tickets from being created.
• Dashboards show how emails are classified, how that compares with agents' manual changes, and how the model performs in each region.

‍

Result & Next Steps

Every Vandemoortele region now turns customer emails into cases the same way, with orders and service requests separated from clutter automatically. Vandemoortele uses Tekst Process Intelligence to explore new automation use cases across its customer service & quote-to-cash operations.

• 90% classification accuracy in testing, across orders, service requests and clutter
• One classification standard for every region
• Consistent, accurate ticket creation, with contacts matched to the right customer
• Clear separation of genuine customer requests from clutter and internal mail
• Ongoing visibility of model performance in every market

‍

‍

‍