Allstad manages church buildings and properties across the country. As custodians of the parsonage heritage, they have a responsibility to both history and the future. Today they sit on enormous volumes of valuable data. The challenge was that the information was spread across systems such as Asta, ACOS’s WebSak, Facilit and various Microsoft-based solutions.
The question was both fundamental and complex: how can they use artificial intelligence to make everyday work more efficient – without either losing control or locking themselves to specific systems?
We wanted to test whether AI could help us get more out of the data we already have, not least from our large document archive containing several hundred years of history.
A pilot that can carry production
Allstad decided to get started with a pilot project with TellusR. Instead of locking themselves to a specific system vendor, they got a test setup where data from several different systems was connected and made searchable, and Allstad’s first AI assistant was configured as a deliverable from the PoC. That made data from several different systems available in a single interface – through AI search and an AI assistant.
The threshold for testing was low, but the foundation was built for the future. The architecture established in the pilot was the exact same one that would carry the solution in full production.
On top of that foundation Allstad can configure AI assistants and agents as they wish, safely connected to the data they choose.
The low threshold to get started means a lot to us. In just a few weeks we saw concrete results. We can now search across our systems, and it gives us a much simpler basis for figuring out how we can best use AI going forward.
Transition-ready, with room for manoeuvre
Allstad chose to continue after the PoC (proof of concept), and the solution is now in production. They are building on exactly the same architecture that was put in place as part of the test. By placing TellusR as an independent layer outside the line-of-business systems, Allstad achieved what many organisations struggle with when they implement AI: real technological room for manoeuvre.
In practice this means they are free to replace both underlying systems and AI models later. They are transition-ready.
With this approach, Allstad is not only preserving the past. They are also exploring how new technology can help manage cultural heritage more intelligently in the future.
Our experience shows that for many Norwegian organisations, getting started with AI is the biggest barrier. That is why we emphasise quick, simple pilot projects that deliver learning from day one.
The difference is that even though the threshold for testing is low, the robust architecture is still in place if they choose to go further after the PoC. That lets customers like Allstad test safely, and keep their technological room for manoeuvre when the solution goes into production.
