It takes a strong core to build something powerful. Let’s head into the world of tech and experience the three acts of Fusion to find out why.

Before we head deeper down into the Fusion rabbithole, we’re going to hop into the time machine to find out how it all began. Originally, Trond Karlsen and Jannicke Sangolt were asked to take data generated in a project completion tool, create a web front and display the info there. 

“But we found that to be too old-fashioned and wouldn’t really generate any real value,” Trond says.

“So instead we locked ourselves into a meeting room for a couple of days, played around with post-it notes and came up with the idea,” he laughs.

After getting a go-ahead to create a proof of concept, Trond and Jannicke got to work. They spent 4-5 weeks on proving that their idea would work and when they presented their results the ball started rolling and they established a team.

Trond explains that one of the biggest problems they wanted to solve from the start was handling users and their access - an extremely time-consuming task that they hope to be able to solve in the future.

Hans is one of two from the original Fusion team, and has been working on the project since 2016. One of the challenges they faced was that data is spread across many tools and each tool has a different metrics to define a project, Hans explains: 

“We can’t just start with a blank slate and force everyone over to a new system, we have to make use of the data we already have. And the more we work our way through the system, the more we see that data is made by a system - for that system.”

Data sets would grow so big that web server processes would run out of memory when processing them.

Another challenge was a long load time, which we solved by implementing “data factories” and delta queries - allowing them to do partial reads of the source. 

“Basically, we have snapshots from nightly caching and when users access certain data we can run a request to see if there’s been changes to the data,” Hans explains.

“These two approaches leads to current data with a reasonable wait time without servers buckling,” he adds.

This information was usually based on spreadsheets and overviews that said who’s who, where they worked and what their job was. 

“When we managed to digitize this data it was a goldmine of information to use for all of Fusion,” Hans says.

By updating user information in the people component it’s automatically updated across the board. There’s also a history to users so we can go back in time to see who worked on what and when - which Norwegian regulations require us to do. 

“We had some trouble in the beginning, explaining to people who just wanted a web-version of the presentations that there’s all kinds of functionality to it, but we got there in the end,” Hans laughs.

There’s no way we can try to tell the story of Fusion without mentioning the app teams. Like you might remember, Fusion is more of a framework for the apps to run on. 

“The most exciting thing is that we’re using the latest tech available. Using React and the component-approach means we can develop apps faster, while just shopping for components in the “Fusion shop”,” developer Hallgeir Gutvik says.

Each app is separate from each other but they still talk together and with the core components: affectionately known as microservices. 

Having a core portal like Fusion in place means that spinning up a new app is incredibly easy - all it takes is a command line and off you go. And while the biggest challenges can be a pain they’re also the most exciting part.