By Ved Antani, SVP Engineering AI & ML & MD India, New Relic
At its inception, cloud computing promised to transform how businesses access, manage, and deploy technologies. The initial allure of cost efficiency and scalability were attractive propositions for businesses world over and were central to the advent of modern IT strategies. As cloud computing matured, rising costs have clouded its value proposition. The rush to adopt AI and ML technologies, coupled with complex pricing models, has often led to bill shock.
These costs aren’t going away. Gartner predicts that cloud spending will jump 21.5% in 2025 and that cloud infrastructure and platform services will be the largest category of enterprise cloud spend. Businesses are under increasing pressure to control cloud costs and maximize value from their multi-cloud environments, where data volumes are exploding with the use of AI.
In addition, some cloud providers have complex pricing structures, and in some cases, customers are charged for volumes that aren’t automatically removed for data transfers between cloud services and regions. Any lack of visibility across data pipelines in the cloud can result in overspending, misallocated resources, and unpredictable future costs.
Finding wasted resources, optimizing costs
In the age of virtual machines, teams often use oversized and underutilized resources, especially with Kubernetes-based container computing. It is common for teams to set very high request resources, well above the actual demand of workloads, leading to wastage of cloud instances that may be paid hourly, resulting in higher costs. In large enterprises, where multiple teams manage vast amounts of data and cloud instances, overprovisioned resources can go unnoticed. This is where intelligent observability can help. Observability platforms with in-built cloud cost insight and intelligence capabilities offer real-time, comprehensive visibility into cloud resource costs, and can run cost estimates for planned resources. These platforms also offer actionable insights that help businesses align their objectives with estimated costs, while simultaneously alerting relevant teams on unusual spend patterns that may require further investigation. This means both business and engineering teams can predict dynamic cloud spending, and scale back costly overprovisioned resources.
Furthermore, observability platforms with integrated Cloud Data Pipelines (CDPs) can help enterprises build analytics quickly by automating ingestion and data processing workflows, while enabling businesses to leverage new data sources, and support new business requirements. Intelligent observability tools also route, filter, and transform all telemetry data in flight, enabling businesses to maximise the value of data ingestion. This means businesses only pay for valuable data, and ensures that resources aren’t wasted on poor data preparation. On the security side, these tools can run pipelines in the businesses’ own infrastructure to mask and obfuscate sensitive data to ensure compliance.
Having the right data and dashboards that visualise the actions required is paramount for optimising a multi-cloud ecosystem. It speeds up the way enterprises use the cloud and drives better functionality. What was previously a complex conversation around finances is easily translated to actionable insights with intelligent observability.
Tracking and forecasting cloud spend is essential to assessing the demand. A recent study by 451 Research found that 56% of savings came from the correct use of cloud financial tools. By investing in intelligent observability, enterprises will not only give their teams the best experience, they will also make their business more resilient and protected, resulting in better ROI on cloud investments.