Send AI
industry insights

How Melkweg|Fritom Transformed Specialized Knowledge Into Intelligent Automation

By Hylke de Vries - Customer Success Engineer at Send AI

Melkweg|Fritom identified an opportunity to scale their email-based order entry operations. The majority of their customers send transport orders via email. After a 1-month pilot, Melkweg|Fritom implemented Send AI's intelligent document processing platform, automating 70% of their email-based customers, representing approximately 90% of total email orders. The system achieves an 85% straight-through processing rate, enabling the team to focus on high-value customer service. The solution is now expanding across the Fritom Group.

Key Results

  • 90% of email orders automated
  • 85% straight-through processing
  • 24/7 continuous order processing

Company Info

  • Industry: Liquid Bulk Transport
  • Part of: Fritom Group
  • Location: Netherlands
  • Use Case: Transport Order automation
  • Integration: Transpas TMS (Art Systems)
  • Opportunity: Expanding across the group
Truck_MelkwegFritom.jpg

Liquid bulk road transport - Source: Melkweg|Fritom

Understanding Transport Order Entry

Melkweg|Fritom specializes in liquid bulk transport, moving everything from foodgrade, feedgrade and technical liquid bulk products all across Europe and beyond. It's complex, regulated work that requires precision at every step.

At the heart of their operations sits order entry: processing customer transport requests by receiving, validating, and entering them into their TMS (Transpas). Each order contains pickup and delivery locations, product details, dates, references, and special instructions like temperature requirements.

The majority of Melkweg|Fritom's customers send these orders via email, where information is scattered throughout lengthy threads as details are discussed and confirmed. These orders arrive in various formats: email, PDFs, Excel files, or combinations depending on the customer.

The Challenge: Adding Customer-specific Knowledge to the Workflow

As Melkweg|Fritom's business grew, their order entry process presented an opportunity to work smarter. Order information flowed into two primary inboxes where a team of 4 people worked to extract the orders and manually enter them into Transpas.

Order entry wasn't just data entry, it required specialized expertise. Team members like Naomi, who had mastered this role over years, knew exactly how to handle complex scenarios. Most importantly, they knew how to add loading locations and times when customers didn't specify them, and how to calculate delivery dates based on loading dates and transit times. This expertise was valuable, but the team's capacity was stretched thin.

Training new employees took months due to client-specific knowledge that was hard to transfer. Beyond order entry, the customer service team handled many other critical responsibilities. They handled urgent situations like production stops and delays, requiring immediate problem-solving and customer communication. They also managed daily operations, answering questions about loading and delivery times, booking timeslots, and checking transport availability.

These high-value customer service tasks are critical to operations, but difficult to combine with continuous order entry during busy periods. Melkweg|Fritom recognized this was the perfect time to explore intelligent automation for order entry, freeing their expert team to focus on work that truly required their specialized skills.

Liquidtank-Port_MelkwegFritom.jpg

Liquid bulk tank - Source: Melkweg|Fritom

The Qualifications of a Fitting Solution

When evaluating automation solutions, Melkweg|Fritom knew exactly what they needed:

  • High accuracy text extraction: The solution needed to reliably extract order information from various formats regardless of how customers formatted their orders
  • Robust exception handling: Not every order would be perfect. The system needed configurable business rules that could validate data, flag missing information, and apply customer-specific logic
  • User-friendly interface: The team needed transparency in how data was processed, with an intuitive interface they could use immediately
  • Seamless integration: The solution had to work smoothly with Transpas via Fritom's internal service infrastructure
"We're always looking for innovative ways to improve our customer service. We needed a partner who could move fast but also understand the complexity of our existing infrastructure. Send AI proved they could do both."

Christiaan, Improvement Manager & ICT Coordinator, Melkweg|Fritom

From Pilot to Production

With an ambitious timeline, Melkweg|Fritom made sure success was evident. The pilot was very important, but of course moving to production was where the real value lives.

1. The Pilot Phase

During the pilot the Send AI team worked closely with Sabine from order entry to test the platform on a subset of customers. The intuitive, email-like interface meant the team did not need a lengthy onboarding. By the pilot's end, during a presentation attended by both Melkweg|Fritom and other Fritom Group companies, the results were compelling. The decision to move forward followed soon after.

2. Implementation

With Naomi leading the operations side, Melkweg|Fritom moved into full implementation. The team systematically expanded automation customer by customer, day by day, capturing their specialized expertise and making it accessible to everyone.

3. How It Works Today

Order entry employees forward relevant emails to Send AI or drag-and-drop them directly from their inbox. For the largest customers, automatic forwarding rules handle this seamlessly. Within seconds, Send AI extracts order information and enriches the data based on the business rules . For orders that pass validation (approximately 85%), data flows automatically to Transpas. When orders don't pass validation, the team can resolve issues in just a few clicks, which in turn makes the system smarter.

Naomi_MelkwegFritom-1.jpg

Naomi working with Send AI - Source: Melkweg|Fritom

"Every week we add more customers to automatic processing, and every week I see my team getting more breathing room. We're now focusing on the complex work that truly benefits from our expertise, solving problems and providing excellent customer service."

Naomi, Order Entry & Customer Service, Melkweg|Fritom

The Results: Strategic Control & Scalability

Key Impact

  • 90% of email orders automated with 70% of customers now processed automatically
  • 85% straight-through processing with no human intervention required
  • Continuous 24/7 processing outside of business hours
  • Significant data quality improvement through consistent validation

Team Transformation

The customer service team now focuses on exception handling and high-value customer service, work where their expertise and personal approach add the most value.

Knowledge Scaling: Business rules encode the expertise that experienced team members like Naomi developed over years. This specialized knowledge is now consistently applied across all orders and accessible to every team member, including new hires.

Control & Transparency: When an order doesn't pass validation, the team can see exactly why and resolve it in just a few clicks. They can create business rules to prevent recurrence. This level of control builds confidence and makes the team more autonomous.

What's Next: Opportunity to Expand Across Fritom Group

The success at Melkweg|Fritom has created an opportunity to expand across the Fritom Group. When representatives from other Fritom Group companies attended Melkweg|Fritom's pilot presentation, they immediately saw the potential for their own operations. Other companies within the Fritom Group have begun working with Send AI on similar automation projects, including transport order automation and CMR document archiving. These projects are showing positive results. Looking ahead, the team is exploring automatic intent classification, enabling the system to understand whether an email contains a new order, an update, or a status inquiry, and take the appropriate action. This capability is expected for early 2026.