Capgemini showcases the power of GenAI at GlasfaserPlus in predicting fiber optic rollout delays

Complex projects like the expansion of fiber optic networks are influenced by a diverse set of mostly external factors, such as municipality permits, approvals, the field worker availability of construction partners, and weather conditions. Unsurprisingly, the expansion of fiber optic networks cannot always be planned months in advance. Delays happen and dealing with them is part of the day-to-day work of telecom companies.

GlasfaserPlus, a joint venture of Deutsche Telekom and IFM Global Infrastructure Fund, specializes in rural network expansion and is on its way to become one of Germany’s leading fiber optic infrastructure companies. When attempting to provide next generation internet-service to nearly the entirety of Germany, the real challenge was the sheer number of parallel projects numbering in the thousands.

Because of this scale, GlasfaserPlus’ back-office receives a massive amount of miscellaneous documents via different channels, including faxes, letters, and emails, as well as unstructured comments from construction workers and legitimate and unjustified claims from municipality mayors or partner companies. Every piece of data requires meticulous reading, classification, and documentation so that the company can use it to calculate construction delays.

To help streamline this process, GlasfaserPlus engaged Capgemini to help identify and introduce an effective solution.

Streamlining delay calculation with Gen AI

After reviewing the company’s needs and existing processes, a small, efficient team of Capgemini intrapreneurs suggested the application of an advanced Gen AI system to handle the data while ensuring compliance with data-privacy and data-protection regulations. Once GlasfaserPlus stakeholders signed off on the plan, the project team began the development process.

“When the business side approached us in IT with the requirements, we did not see a viable solution based on conventional means. Fortunately, our colleagues from Capgemini, who were already working on the project, were able to draw on their expertise and the broad knowledge in Capgemini’s background. Without Capgemini, it would probably have taken much longer to tackle these problems.”

Henk van Es,
Head of IT, Glasfaserplus IT Services GmbH

The result was a solution that uses a Large Language Model (LLM) to convert unstructured data – in the form of PDF documents – into structured data. The system analyzes and evaluates these documents based on well-formulated prompts developed collaboratively by GlasfaserPlus and Capgemini. Throughout the development process, the project team followed an iterative approach while working with GlasfaserPlus’ IT department to respond to specific challenges.

Following this automated review, the Gen AI solution extracts the most important core information and stores it in a predefined, structured format. As a result, GlasfaserPlus employees now gain access to easy-toread data that enables them to rapidly and effectively predict delays to fiber optic expansion projects.

Innovation supported by continual improvement

After eight weeks of close collaboration, GlasfaserPlus and Capgemini released the Gen AI solution, which immediately began delivering business value. Moving forward, the final solution will help to further minimize time-consuming back-office-tasks related to delay management of scaled network-expansion projects. Following the successful development and launch of the solution, GlasfaserPlus and Capgemini are now working to improve the system by making it more robust, adding different features to automate monitoring and reporting, and connecting additional data-sources to a meet even more diverse set of requirements. In doing so, GlasfaserPlus will continue to differentiate itself and define market-leading services within the German telecom market.