Hundreds of hours saved with AI-powered documentation solution

Baneservice, a Norwegian railway contractor, is often required to meet BREEAM certification standards for large infrastructure projects. BREEAM is a comprehensive framework for evaluating environmental performance, covering 248 criteria across eight categories. For Baneservice, the documentation process involved collecting vast amounts of evidence and writing formal compliance descriptions for every single criteria. This repetitive, manual task consumed significant time and required input from both internal experts and external consultants. With sustainability documentation becoming a regulatory and client expectation, Baneservice needed a more efficient and scalable approach.

While initially exploring AI for another project, Baneservice and implementation partner Novacare identified the BREEAM certification process as a perfect use case.

“I saw it as a textbook example of where large language models would provide value. It's a really tedious process that requires reading and processing a lot of texts, which is what large language models are really good at. It's a perfect match.”

- Sigve Fast, AI expert and developer at Novacare, Sigve Fast.

Novacare and Baneservice then began the development of a custom, tailored AI-powered application to support compliance documentation. The solution integrates with SharePoint and leverages large language models (LLMs) to automatically generate draft compliance descriptions based on selected project evidence.

To ensure accuracy, the application includes a human-in-the-loop workflow. For each critera, the AI produces three variations using different models and prompting techniques. BREEAM experts then review, edit and approve the drafts, significantly accelerating their work without compromising quality. The solution was delivered rapidly, with flexible integration and secure data handling.

The AI solution has delivered significant benefits for Baneservice’s BREEAM documentation process. It has reduced hundreds of hours of manual work per project, allowing experts to focus on reviewing high-quality drafts instead of starting from scratch.

“The solution gives the BREEAM experts superpowers because they can accelerate their own expertise by taking some of the manual labor out. That might be the most important result: happy workers”

- Alexander Guy Grønli-Raastad, Business Development Innovation at Baneservice.

‍

This shift has not only improved the work experience but also lowered costs by reducing reliance on external consultants during the documentation phase.

“We save time and can cut down costs on hiring additional resources and experts from consultancy firms, and that is money saved straight on the bottom line”

The AI-generated texts are based on structured evidence and document references, ensuring greater accuracy and traceability. The underlying architecture is designed for scalability, making it easy to extend the approach to other document-heavy processes across the organization. Furthermore, the application was built and deployed in a fraction of the time traditional development would have required.

• Time savings
• Improved work experience
• Reduced cost
• Increased accuracy and traceability
• Scalable impact
• Fast implementation