Overview

Fast scaling can be considered an issue when it starts impacting your budget through uncontrolled cloud spending. In one particular instance, Star Systems helped a rapidly scaling SaaS company address the most important concern that they were experiencing: their cloud spending kept on growing from month to month, yet they could not track where the money went. A crucial consideration made when designing the system was its capability to manage a high volume of claims at once, without errors, and with effective control measures.

Organization’s AWS architecture evolved organically as the product and client base expanded. However, this resulted in inefficiencies that were not immediately apparent.

• Delays in claims processing negatively impacted customers’ experience
• High costs of operations owing to high manual effort
• Inability to detect fraudulent claims on time
• Scalability issues that made processing high volumes of claims difficult

Approach

We used a targeted Cloud FinOps audit lasting two weeks to address the problem, ensuring the identification of inefficiencies and the execution of necessary improvements without affecting existing processes.

Firstly, cost visibility had to be achieved and we created a consistent tagging strategy, ensuring all resources were tagged according to their respective teams, departments, and environments. This, alone, helped management gain clarity regarding where cloud costs were going.

Then, efforts were made to optimize the resource usage. Through an analysis of usage patterns, Star Systems was able to determine which instances were underused and oversized. Around 60% of the compute resources were rightsized.

The second step was resource optimization. Discovering there were many instances with underused resources and instances that were oversized. Around 60% of their instances were optimized, matching their resource capacity to their actual requirements.

To ensure long-term savings, a 1-year reserved instance policy was put in place for predictable loads. This helped save money on computing resources while not compromising performance.

At the same time, automatic procedures were enabled to detect and shut down any idle resources, making sure that inactive services would not incur additional costs.

Finally, a custom FinOps dashboard was developed to give real-time insights into cloud utilization and cost savings potential.

Impact on Business

The effects were fast and profound. In just a short time, the organization realized huge savings on its cloud computing expenses:

• From $180,000 per month down to $112,000 per month
• Saving 38% with no compromise in efficiency or scale
• Saving $816,000 per year annually
• Gaining continuous visibility with the help of a FinOps dashboard

But perhaps more importantly, it shifted from being reactive with cloud costs to proactively embracing a FinOps mindset.