By Sergey Merkelov
Infra R&D Director at Priority Software

Read More >

 I’m the Infra R&D Director at Priority Software. Our main product is an ERP, widely used around the world. In the last few years, we added more lines of business, like retail, education, and hospitality.

My department is building a standardized cloud platform that serves all of these businesses. Our AWS organization counts multiple accounts, split by line of business and budgets.

Most of our workloads are containerized on Amazon EKS. We operate multiple clusters across four AWS regions, each with a significant number of nodes.

Savings Plans were the right model, but the problem was keeping high coverage

Savings Plans have always been a good fit for us because the infrastructure keeps evolving. Usage grows continuously, instance types change, and new generations of machines are released, so flexibility matters.

For me, the main challenge was keeping high coverage over time.

Savings Plans were managed through a partner. They were usually updated once a month, in a half-manual process. My main goal was maximum coverage, but the pattern was always the same: coverage looked good right after an update, then gradually dropped as workloads shifted. You might start at 90%, then two weeks later, you’re at 85%, then 82%, and by the end of the month, it can reach 80%.

With our setup and continuous growth, manual management simply couldn’t keep up. When a human is involved in a process, there’s always a gap.

Moving to automation: what mattered most was the people

When we heard about Zesty, it was clear this was the right solution for us.

The onboarding itself was very simple. We connected it and did the math. But the bigger part was the transition plan. We needed a gradual rollout to release existing commitments and ramp up new ones safely.

The Zesty team handled it end-to-end, including communication with our partner. It was full service. They were very professional and great to work with. I had a lot of questions along the way, and they patiently answered everything.

They also handled the complicated calculations. There are a lot of nuances here that most people aren’t aware of. Workloads aren’t the same across the week, weekends are lower than weekdays, and you need to factor that into coverage. They helped us figure out the best fit, including the combination of 1-year and 3-year commitments and the flexibility over time.

It’s not enough to have good software. You also need good people behind it, and that was definitely the case with Zesty.

The impact: full coverage without the manual work

We increased coverage from around 85% to 99%. With a large number of workloads, that difference adds up to real savings.

The biggest change is that we can stay on that very high coverage all the time. Savings plans adjust automatically as we grow day to day. That’s really the main value. And if usage ever needs to scale down in the future, daily micro Savings Plans make it much simpler to adapt.

I still look at FinOps data every day or two, sometimes daily, but now it’s just a quick check to confirm coverage is high. It’s five minutes and moving on.

For us, the value is simple: it used to be manual and always chasing the tail of growing workloads. Now it’s automatic. That’s the win.

What I would like to see next

Today, our main focus is compute, but we also have databases like RDS, even if it’s more marginal, with basically zero coverage. I’m definitely waiting for Commitment Manager to optimize it as well.