When the Cloud Starts Running the Show: How Apollo Took Back Control
Apollo Group has come a long way from its roots as a bookstore, first opened in Tallinn in 2000. Since then, the company has expanded into cinemas and restaurants, growing into a leading entertainment operator in the Baltics. Today, the group employs over 4,000 people, serves more than 1 million club members, and generates a net profit of €3.8 million, while pursuing ambitious expansion in Finland. Behind the scenes, that means an IT environment where downtime isn’t really an option and where systems need to scale as quickly as the business does.
For Apollo’s Head of IT, Vsevolod Boikov, the cloud is less about technology and more about discipline.
“At one point, we realized we didn’t fully understand what was going on anymore,” he says. “The infrastructure was growing, but the benefits weren’t keeping up. Costs told the story. To maintain the same level of service, the team kept adding resources — even though the number of users hadn’t changed. Nothing dramatic at first. Just an extra €500 a month. Then another. And another. Until hosting alone was costing close to €50,000 per month — a level we find hard to justify.”
The turning point came with a closer look at their cloud architecture — and a conscious effort to build in-house expertise. Within a year, Apollo had cut its cloud costs nearly in half.
But cost wasn’t the only issue. Reliability was, too. A single technical failure or DDoS attack could take the entire service offline globally. The fix was structural: systems were split and segmented so that one failure wouldn’t bring everything down with it.
Today, Apollo’s stance on the public cloud is firmly positive. Microsoft Azure, in particular, has become a core part of the company’s infrastructure, enabling both scale and speed. But the mindset has shifted.
If the early promise of the cloud was speed, the reality, Boikov says, is control.
“Without actively managing it, you lose visibility, and once that’s gone, efficiency follows.”
That shift has also changed how the team works. Cloud management is no longer a shared responsibility across development teams; it now sits with a dedicated DevOps function. According to Boikov, the impact has been clear across the board — from cost to performance to risk.
Even so, cost control remains the hardest part.
“It’s very easy to add resources, what’s much harder is understanding whether you actually need them.”
Vsevolod Boikov, Head of IT, Apollo Group
Vsevolod Boikov, Head of IT, Apollo Group
That tension — between performance and cost — doesn’t go away. It becomes part of the day-to-day.
The cloud offers flexibility and speed, but it doesn’t remove responsibility. Some things, like scalability, are best left to the cloud. But architecture and cost discipline, Boikov argues, can’t be outsourced.
Not everything belongs in the cloud anyway: some systems like HR or finance, for example are still simpler and cheaper to run on-premises. Legacy systems and licensing constraints can also make full migration impractical.
Apollo’s experience is a reminder that moving to the cloud is not the same as mastering it.
“The cloud gives you tools,” Boikov says. “It doesn’t necessarily solve all your problems.”
The real work lies in finding the right balance — and having the capability to manage it.