Offering software-as-a-service (SaaS) human resources and financial management services to around one million European customers, Norwegian software firm Visma places a great deal of importance on delivering new features as rapidly as possible in order to meet customer needs.
However, its on-premises data center infrastructure presented a major obstacle to this goal. According to Visma’s Chief Cloud Architect, Alexander Lystad: “With a data center model, we spent a lot of time managing infrastructure, and our developers lacked the agility and scalability they needed. For many of our products, we were releasing new software features six times a year, which isn’t enough to be competitive. We wanted to give our developers self-service capabilities to change that.”
To enable this transformation, Visma migrated many of its workloads to Amazon Web Services (AWS), with Lystad saying that “AWS offered the agility and scalability we were looking for”.
Visma began with its .NET-based and SQL Server–based applications, which it migrated to Amazon Aurora MySQL, with Amazon EC2 used for on-demand compute capacity.
This migration involved the company building Amazon Machine Images (AMIs) to achieve automatic scalability, with Amazon Relational Database Service (Amazon RDS) serving as the database, while AWS CloudFormation templates allowed Visma to quickly provision resources. As Lystad says, this enabled a more seamless and cost-effective migration.
The migration has seen Visma reduce the development time it needs to build and deliver software, going from six software releases a year, to more than six per month. The company can also update software during office hours without downtime, increasing the reliability of its service.
The company has also reduced deployment time for employee management applications from 30 minutes to 5, while it now only takes on average 30 seconds for the company to scale up an application, compared to 5 minutes with its previous environment.
Lystad says: “Using AWS, we have been able to consistently reduce our average cost per user for one of our systems by 50 per cent while increasing service quality and growing our business by four times since starting the migration”.