How Hapag-Lloyd uses Amazon Bedrock to transform customer feedback into actionable insights

Hapag-Lloyd, a global shipping leader operating 313 container ships and a container capacity of 3.7M TEU with 14,000 employees across 400+ offices, previously relied on manual customer feedback analysis. Product managers exported CSV files every two weeks, read hundreds of comments, and manually categorized sentiment and themes — a process that could take hours or even days. To scale and accelerate decision-making, the Digital Customer Experience and Engineering team built an AI-native solution using Amazon Bedrock, Elasticsearch, and open-source frameworks like LangChain and LangGraph. The architecture leverages AWS services: Lambda functions ingest feedback data from Amazon S3, Amazon Bedrock (with models from Anthropic, Meta, and others) extracts sentiment and themes, and indexed results in Elasticsearch. Stakeholders access insights via Amazon ECS, with automated email notifications through SES. The entire deployment is managed via AWS CloudFormation for scalability and maintainability.

This generative AI pipeline transforms the feedback loop by automatically collecting customer comments, summarizing sentiment, and identifying recurring themes. Product teams now focus on strategy and innovation instead of repetitive data crunching, enabling faster, smarter product decisions. Hapag-Lloyd's move toward becoming AI-native demonstrates how even traditional industries can harness foundation models to turn operational data into actionable business insights at scale. The solution is production-ready, secure, and privacy-compliant, built on Amazon Bedrock's fully managed service with a choice of high-performing models.