About customer

Kar-Tel, operating under the Beeline Kazakhstan brand, is one of the country’s largest telecommunications providers, serving over 10 million customers with mobile, internet, and digital services. As part of its operations, the company manages a large fleet of virtual machines (VMs) hosted in Microsoft Azure.

Due to the scale and complexity of this infrastructure, many VMs were often provisioned with surplus resources such as extra CPU cores. This led to unnecessary cloud expenditure and required IT specialists to spend significant time manually reviewing periodic reports and identifying optimization opportunities. 

Challenges and solutions

Beeline Kazakhstan set out to address two key challenges:

• reducing unnecessary cloud costs caused by over-provisioned resources, and

• freeing IT teams from the burden of repetitive manual analysis.

The company’s objective was to establish an automated, scalable process that could not only collect raw VM data but also process it into actionable insights. This would enable IT specialists to quickly identify which VMs could be resized to optimize tariff plans and reduce expenses — all while ensuring smooth day-to-day operations.

Beeline Kazakhstan partnered with Awara IT to design and implement a corporate solution powered by Microsoft Power Platform and AI capabilities. The platform was fully integrated into the existing Microsoft 365 ecosystem and Microsoft Teams, providing IT staff with a user-friendly and accessible tool for continuous resource optimization.

Key elements of the solution included:

• Automated upload and processing of VM configuration reports, with AI components interpreting unstandardized data.
• Analysis logic highlighting the top 10 VMs with the highest downsizing potential, based on underutilized CPU cores and other parameters.
• Structured, actionable reports eliminating the need for manual spreadsheet reviews.
• Integration with Microsoft Teams for easy access and automated notifications.
• Scalability to support dynamic infrastructure and frequent reporting cycles.

While AI supported data interpretation, the calculation and ranking logic was implemented through deterministic Power Platform workflows to ensure consistent and reliable results. 

Results

Within the first six months after launch, the platform delivered measurable outcomes:

• Over 500 optimization reports automatically generated.

• 40% reduction in IT workload related to manual report analysis.

• Improved consistency in monitoring and reduced risk of missing cost-saving opportunities.

• Significant cost reductions through ongoing VM optimization.

• Faster and more transparent decision-making within the IT department.

The solution established a structured, repeatable process that strengthens Beeline Kazakhstan’s operational efficiency and cost control while reinforcing its reputation as a digitally progressive telecom provider.