Customer support · Production

Hapag-Lloyd Uses Blue Prism's Robotic Operating Model to Scale Intelligent Automation

The problem

Hapag-Lloyd needed to implement intelligent automation at scale across more than 130 countries, requiring clear governance, defined roles and responsibilities, and a model that could accommodate both a central COE and remote regional business units.

Workflow diagram · grounded in source
1
Incoming Customer Disputes
trigger
“the dispute capturing digital worker has been integrated with AWS-based machine learning (ML) modules to manage incoming customer queries and dispatch the appropriate responses”
2
Digital Worker Scans Email Inboxes
ai_action
“The digital worker accesses multiple email in-boxes to find all disputes and passes them to the ML module”
3
ML Module Classifies Request Type
ai_action
“the ML module, which determines the request type”
4
Dispute Routed to Staff Member
routing
“Once determined, the digital worker sends the dispute to a staff member.”
Reported outcome

Hapag-Lloyd deployed 50 digital workers across three key business units and saved over 30,000 hours per year through the dispute capturing automation alone, winning the EMEA 2021 ROM Excellence award.

Reported metrics
Hours saved per year (dispute capturing)over 30,000 hours a year
Digital workers deployed50
Additional digital workers planned for production15
Reported stack
Blue PrismAWS
Source
https://www.blueprism.com/resources/case-studies/hapag-lloyd-uses-blue-prisms-robotic-operating-model-to-navigate-intelligent-automation/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Hapag-Lloyd deployed 50 digital workers across three key business units and saved over 30,000 hours per year through the dispute capturing automation alone, winning the EMEA 2021 ROM Excellence award.

What tools did this team use?

Blue Prism, AWS.

What results were reported?

Hours saved per year (dispute capturing): over 30,000 hours a year; Digital workers deployed: 50; Additional digital workers planned for production: 15 (source-reported, not independently verified).

How is this customer support AI workflow structured?

Incoming Customer Disputes → Digital Worker Scans Email Inboxes → ML Module Classifies Request Type → Dispute Routed to Staff Member.

Related customer support cases