Data Engineering for Logistics and Retail
Otto, Kühne + Nagel, EOS — Logistics, Retail & Finance
How Otto and Kühne + Nagel put data platforms to work for operational excellence — and how EOS built a data-driven debt collection system from the ground up.
Real-time shipment data, data quality as the foundation for AI, a new core system for debt collection
Challenge
Three companies, three data challenges:
Otto needed a scalable platform for product data streaming. Kühne + Nagel required real-time processing of shipment data. EOS rebuilt its entire core system for data-driven debt collection from scratch — a three-year project with up to 8 FTE.
Starting Point
- Isolated data silos with no integration at all three clients
- No real-time processing of logistics data
- Product data from dozens of sources with no unified format
- The legacy system at EOS (Fidibus) needed to be replaced by a modern core system
Approach
Otto: Built a streaming platform for product data, developed automated data pipelines, and integrated them with existing systems.
Kühne + Nagel: Developed an event-streaming architecture with Apache Kafka, real-time processing of shipment data, and a central data platform.
EOS: A strategic partnership under an open-ended framework agreement. Over 15 consultants working in cross-functional teams — architects, full-stack developers, DevOps engineers, product owners, and business analysts. A SAFe-based agile process built on Scrum.
The core levers of the project: an analytics engine for self-learning collection processes, process automation, standardized interfaces for third-party integration, and a dynamic resource-planning system.
Solution
Scalable data platforms with real-time processing. At Otto, the focus was on data quality and product data management. At Kühne + Nagel, on event streaming and operational decision support. At EOS, an entirely new core system (Project FX) built on a modern stack: Golang, Kotlin, React, AWS, Kafka, Kubernetes.
Value / Results
- Data available in real time instead of overnight batch processing
- A unified data foundation for AI initiatives
- Reporting time cut from days to minutes
- An analytics engine powering data-driven, self-learning collection processes at EOS
- A scalable architecture for growing data volumes
Discuss a similar project
Tell us about your starting point — we'll show you what the path forward can look like.
Data-Team
„In 30 Minuten wissen Sie, ob Ihre Datenbasis KI-ready ist."
Book a meeting with Data-Team