Best practices to manage data and services on Azure optimizes cloud costs by 25%

Sigmoid evaluated the data workloads, subscriptions, and resource groups on the Azure cloud and implemented best practices such as serverless architecture and autoscaling. In the first month, a 25% reduction in cloud subscription costs across development, QA, and production environments was observed.

Business Challenge

The customer, a multinational producer of health, hygiene, and nutrition products, was facing cost optimization challenges with their Azure cloud environment. More than 200 analytics projects for sales and supply chain were deployed on the cloud environment. However, there were discrepancies in the allocation of resources and instances, as certain projects require more while others less. This resulted in an increased monthly cost of Azure subscription going beyond the allocated budget. The lack of documentation of different resource groups and decentralized decision-makers led to inconsistent cloud usage and optimization.

Sigmoid Solution

We evaluated the cloud platform subscriptions, resource groups, and instances to establish a cost baseline. Then, we created a detailed breakdown of Azure subscription and infrastructure costs for different resource groups. Using this, we collaborated with key stakeholders to develop a plan to optimize costs. The plan included pausing features not in use, enabling rightsizing and serverless architecture, deleting unused resources, and enabling autoscaling. We also implemented cost optimization measures like spot instances, job clusters in production, and auto-termination. By doing this, we helped the client reduce costs while maintaining and improving performance, optimizing resource usage, reducing waste, and improving infrastructure efficiency.

Download case study

Business Impact

An optimized Azure cloud infrastructure established through best practices for setting up production, development, and quality assurance environments, resulted in remarkable savings in costs. Automating the process of scaling and pausing certain services when not in use, increased the efficiency of processes. Also, to maintain track of expenditures, automated email alerts were established that provided daily, weekly, and monthly expense summaries to important stakeholders.