Connecting data and bringing it to life with AWS for Lundbeck

With NNE’s support, Lundbeck increased the value of their existing production data. When the proof of concept delivered by NNE brought immediate value, Lundbeck decided to take the solution even further with a full-scale AWS production lake.

All pharma manufacturing companies want to gain value from their production data. For Lundbeck, the journey began with an on-premise graph database built by one of NNE’s data engineers. Quirkily named “SuperNinja1”, this proof-of-concept database acted as a small data lake that could store, combine, and query data extracted from their Enterprise Resource Planning (ERP) system and Laboratory Information Management System (LIMS).

Lundbeck were impressed at how quickly this simple solution helped reduce scrap from production, which sparked an interest in building a full data-lake on AWS. So, alongside Lundbeck’s Global IT, NNE’s experts built a solution that could pull in all data for deeper production insight.

The system had to be scalable, accessible, and easy to maintain; if end users couldn’t understand the data, it had little value. Thus, the project team also built a custom frontend on the serverless infrastructure to deliver business insights gained from data analysis and modelling.

Long term, this user-friendly interface will allow production staff to check daily for process issues (rather than one laborious process each year for the Process Quality Report). It will also include predictions to optimise processes. For example, by combining weather station data on humidity with typical tablet drying times using filtered outside air, the tool will recommend the optimal level to operators.

Key benefits:

  • Increases data availability by having one central place for collecting data across the organization.
  • Enables new data related questions to be asked and answered, new business cases to be proposed, and new optimizations to be realised.
  • Enables data analysis on data from previously separate sources.
  • One place to run machine learning models and display predictions.
  • Impacts top and bottom line by decreasing scrap and increasing yield – 3 MDKK already saved
  • Anomaly detection expected to save 1-2 MDKK per year
  • Frontend delivers clear insights into data to non-data scientist.
  • Scalable data storage that is cost efficient and easy to maintain.

Quick facts





Project duration


Services provided

Manufacturing intelligence, digitalisation


Amazon Web Services data architecture