Historical information loses its value when it exists but is difficult to find or use in day-to-day work. Many organizations face exactly this challenge. Archives built up over decades, printed reports, and documents stored across different locations contain valuable knowledge. But as long as this information exists only on paper or in individual PDF files, it remains largely inaccessible to modern processes.
Wolfsberger Stadtwerke faced this challenge as well. The company manages a contract archive dating back to 1901. At the same time, it regularly collects water measurement data that, until now, was available only in printed reports or PDF files. While these two use cases are very different in nature, they share the same objective: making information accessible and efficient to use in daily operations.
Two Challenges, One Goal
The first use case focused on digitizing historical contract documents. Property transfers, lease and rental agreements, construction plans, and utility line documentation had been collected over more than 125 years. Finding a specific document required manually searching through physical folders and archives. There was no digital document management system with search capabilities.
The second use case involved drinking water quality. Results from regular water analyses were available only as printed reports or PDF files. As a result, the data could not be analyzed in a structured way or connected with other data sources. Future analyses, such as combining measurement data with weather data from GeoSphere Austria, first required a reliable digital foundation.
Both challenges reflect a situation common to many infrastructure companies and public sector organizations. Digitization often starts with capturing information. The real value, however, comes from turning that information into usable knowledge.
Why Scanning Alone Is Not Enough
Digitizing documents does not automatically make their content usable.
For the water measurement data, accuracy was critical. Even a single incorrectly recognized character or number could lead to incorrect conclusions. The automated extraction process therefore had to be reliable, with results that could be verified.
The historical contract documents presented a different challenge. Many were written in Kurrent script or highly stylized handwriting. They also included different document types, construction plans, and utility line documentation from various decades. Traditional OCR methods quickly reach their limits with this type of material. The goal was not simply to extract fragments of text, but to identify structured information such as the year an agreement was signed, cadastral numbers, or the people involved.
One AWS Data Platform for Two Different Use Cases
Together with Wolfsberger Stadtwerke, EBCONT developed a serverless, NIS2-compliant AWS data platform that supports both requirements on a shared technical foundation.
AWS S3 serves as the central data lake for documents and data assets. AWS Lambda handles automated information processing. Structured data is stored in Amazon RDS for PostgreSQL, providing the foundation for further analysis.
For the water measurement data, EBCONT uses Amazon Textract. The service automatically extracts measurement values from reports covering the years 2011 through 2024. The results are then validated against a benchmark dataset to ensure high data quality.
For the historical contract documents, Large Language Models accessed through Amazon Bedrock are used to help interpret content, extract structured information, and make historical documents, construction plans, and utility line documentation digitally accessible.
Use Case 1: Water Measurement Data as a Foundation for Informed Analysis
For the first time, the platform makes water measurement data available in a structured format. Instead of manually transferring information from individual reports, relevant values are extracted automatically and validated for quality.
A proof of concept was also developed to connect the measurement data with weather data from GeoSphere Austria. This creates a reliable foundation for future analyses and potential predictive models.
Use Case 2: Turning a Historical Archive into a Digital Knowledge Base
More than 60,000 pages of documents are being digitally processed and made accessible in a structured way for the first time. These include historical contracts, construction plans, and utility line documentation spanning more than a century.
The benefits extend far beyond digitization itself.
Until now, there was no complete overview of the information contained in the historical archive. Structuring the content will make it possible to assess documents more systematically and identify relevant information more easily. This creates greater transparency and provides the foundation for systematically reviewing previously unknown contract records.
The next phase of the project will focus on expanding the solution for production use. Planned features include a document dashboard, global full-text search, OpenSearch-based classifications, and preparing the data for the potential use of chat agents.
From Documents and Data to a Digital Knowledge Base
The project has made more than 60,000 pages of historical documents accessible and, for the first time, structured water measurement data from 2011 through 2024 in a digital format. This reduces manual effort, creates transparency across information accumulated over decades, and provides a foundation for future automation and informed decision-making.
Christian Schimik, Managing Director of Wolfsberger Stadtwerke, describes the value of the project as follows:
“The goal of digitizing these documents is not only to enable targeted searches and prepare for the future, but also to understand the past and make it usable by structuring the information contained in more than 60,000 pages of documents.”
Digitization Does Not End with Capturing Information
Digitization does not end with scanning documents or capturing data. Its greatest value emerges when information is structured and transformed into knowledge that supports people in their daily work.
This is the foundation Wolfsberger Stadtwerke and EBCONT have created together. A historical archive built up over more than 125 years and dispersed water measurement data are gradually becoming a digital knowledge base that simplifies processes and supports informed decision-making.
Whether you’re working with a historical archive, document management, or a data platform, Harald Kerck is available for an initial, no-obligation conversation to explore how you can make your information assets more accessible and easier to use digitally.