BAMA Gruppen AS is one of Europe’s leading companies in fruit and vegetables, with operations across Norway and international markets. Headquartered in Oslo, the company operates across grocery, food service, convenience retailing, industry and flowers, supported by a supplier network spanning more than 80 countries. 

At this scale, valuable business knowledge is spread across reports, presentations and internal documents. Making that knowledge easy to access when employees actually need it becomes a challenge of its own. 

Within BAMA’s marketing department, getting the right information could mean searching through internal resources or asking the person who knew where to find it. 

With numerous questions repeatedly reaching the same person, BAMA needed a more efficient way for employees to access existing company knowledge independently. 

The objective was clear: make internal knowledge easier to reach without relying on one person to act as its gateway. 

But the requirement went beyond implementing a generic AI chatbot. BAMA needed a solution that could work with its internal sources, make AI-generated answers traceable and support the broader information workflows of its marketing team. 

Fabres designed and developed a custom AI-powered Insights Portal tailored to BAMA’s internal workflow. 

At its core, Fabres implemented an AI assistant that enables employees to ask questions in natural language and receive answers based on BAMA’s internal documents. 

Instead of navigating multiple resources or asking a colleague where to find information, users get a much shorter route: 

Question → Answer → Source 

Fabres built source transparency directly into the experience. Users can see the internal materials behind an answer and open referenced documents directly within the portal, making it easier to verify information and move forward without switching between multiple tools. 

Fabres also extended the platform beyond AI-powered Q&A with: 

• Knowledge resources – frequently used reports, documents and links organized in one place. 

• Built-in feedback – users can rate individual answers and entire conversations, creating valuable input for improving the AI experience. 

• Event management – employees can discover events and webinars, access registration links and add selected events directly to Outlook. 

• Configurable dashboards – different content and dashboards can be prepared for different internal groups. 

• Admin control – administrators can manage portal content and rearrange its elements as internal needs evolve. 

Because the business challenge was bigger than “we need a chatbot.” 

Standard AI tools could address part of BAMA’s use case. Fabres built around the bigger picture: how employees actually find, verify and use internal knowledge in their daily work. 

That meant combining AI-powered access to company information with source transparency, document previews, feedback mechanisms, events, knowledge resources and configurable dashboards within one dedicated environment. 

Instead of forcing BAMA’s workflow into the limitations of a predefined product, Fabres built the technology around the workflow. 

At the time of the project interview, the Insights Portal was operational. 

BAMA employees could use one environment to ask questions based on internal documents, review the sources behind AI-generated answers, open referenced materials and access relevant knowledge resources. 

Fabres helped move the internal knowledge experience away from: 

“Who should I ask?” 

towards: 

“What do I need to know?” 

The solution was delivered by a two-person Fabres development team, working closely with BAMA’s data scientist responsible for the AI component. 

Technology: Azure AI Foundry + custom-developed portal.

Fabres turned AI into a practical tool for accessing BAMA’s internal knowledge – integrated into the way employees actually work. 

The Insights Portal gives BAMA: 

• Faster access to internal knowledge – employees can go directly from a question to an answer and its source, reducing unnecessary searching across documents and resources. 

• Less dependency on individual employees – recurring questions no longer have to rely exclusively on the person who knows where the information is stored. 

• A more organized knowledge experience – reports, documents, links, events and AI-powered Q&A are brought together in one dedicated environment. 

• Less friction in everyday information search – users can find, verify and open relevant information without moving between multiple tools and asking colleagues where to look. 

• AI with a clear business purpose – rather than introducing AI as a standalone experiment, Fabres embedded it into a real operational workflow: helping employees access and use company knowledge. 

• A foundation that can evolve with the business – configurable dashboards, admin controls and a modular approach allow the portal to support different audiences and future requirements. 

Before we started this project we made measurements of how easy it was for the target group to find and make use of relevant insights – both qualitative and quantitative (interviews and surveys). The data showed that satisfaction was low, and getting insights was time consuming. 

An MVP for the insights portal was released late May 2026, and throughout summer we made many iterations based on early feedback from our first users. 

Late August same year, we carried out new interviews and a new survey. The data showed increased satisfaction. They would recommend the product to their colleagues and they disliked the thought of losing this tool. 

This is an MVP, and we now have a super foundation for further development and making the product even better. 

The cooperation with Marcin and Szymon has worked really well. They work fast and always have good ideas about how to do things and looking at the next steps. They take responsibility and have a proactive approach to their work. 

Marcin and Szymon are curious about all aspects of product development, and are therefore good sparring partners about technical issues, user and business value, usability and feasibility. 

We at BAMA like working with them! 

Håkon von Hirsch (product lead at BAMA Tek)