Customers / Akamai case study

Comprehensive Cloud Cost Visibility with Zero Tagging

Akamai

A global leader in cloud services and digital innovation, helps over 100K businesses build, secure, and scale their applications in the cloud. Akamai offers various cloud-based PaaS and SaaS services, such as CDN, edge computing, web and application performance, security, and enterprise solutions.

About Akamai

To stay at the forefront of innovation and provide their customers with latest technological advancements, Akamai continuously introduces new capabilities and products to the market.

Their latest product launch was their SaaS API security platform, which provides full visibility across their clients’ API estates through continuous discovery and monitoring.

It conducts a risk audit of every discovered API, identifies common vulnerabilities, and uses behavioral analytics to detect threats and logic abuse within this fast-growing attack surface.

The Challenge: Unlocking Impactful Data and Business Insights From The Cloud

To support Akamai’s processing needs, its architecture leverages cutting-edge technologies, including Kafka for event streaming, Apache Flink for data processing, ClickHouse databases for scalable querying, an EMR-based data pipeline, and Kubernetes for orchestrating containerized applications.

To drive cost-efficiency throughout their robust cloud infrastructure and maximize the cloud’s revenue potential to the fullest, the engineering team at Akamai knew they had to take their FinOps strategy beyond right-sizing and cost savings.

The FinOps initiatives they were looking to promote started with aligning their technology with business objectives.
Due to their business nature and large and complex cloud environments, gaining end-to-end visibility and tying cloud costs directly to business outcomes became an organizational challenge.

Challenges Faced

Team Bandwidth – Reaching complete cost visibility required many man-hours in tagging every resource and developing internal enrichment tools for cost allocation.

Shared Resources Cost – Tracking actual cloud usage and costs on shared resources was difficult and often inaccurate.

Budgeting and Resource Allocation – The complexity of achieving true cost visibility within the data pipeline makes projecting usage and budgeting cloud environments very challenging and inaccurate.

Overhauling existing systems – Akamai needed to reduce costs effectively without interrupting the business.

“Eliminating the need to tag thousands of resources has freed up my team and we’ve invested our efforts in enhancing our platform significantly.”

Ziv Sivan VP of Engineering

Solution Discovery

Before adopting Attribute, Akamai explored several avenues, including internal tool development and experimenting with various market solutions. However, contextualizing and perfectly attributing cloud spend was a slow, long, manual, and often incomplete process due to the technological limitations of existing solutions in the market today.
They lack a deeper and more granular level of cost attribution, causing blind spots, and require much effort in precisely tagging and logically grouping every resource according to the business metrics that matter.

Their search for a solution that could provide automated, in-depth service-level visibility and clear cost allocation ended when they were introduced to Attribute.

Attribute’s promise of opening the FinOps Black Box and contextualizing every cloud spend, driven by their innovative eBPF technology, offered a new level of cost attribution that required no tags or additional engineering efforts.

Phase I Result: Seamless and Complete Cost-allocation, with Zero Tags.

Integrating Attribute into Akamai’s cloud environments was an easy and seamless process. The platform automatically ingested, categorized, and organized data from multiple sources into application-level groupings, requiring no manual effort from their teams.

They instantly obtained detailed insights into the real-time costs of every service within their infrastructure. Attribute enabled a clear understanding of the cost drivers within their EMR-based data pipeline, revealing how Spark jobs affect compute, database, and data transfer costs.

This financial traceability in cost allocation of shared resources allowed them to quickly identify and adjust the costs of inefficient services and applications.

"Attribute not only automated and simplified the allocation of our cloud costs, it revolutionized it.”

Ziv Sivan VP of Engineering

Phase II Results: Seamless and Complete Cost-allocation, with Zero Tags.

After reaching their desired cost-visibility posture of making every dollar they spend in the cloud accountable, Akamai is currently in the process of making their cloud spend more impactful, from a business perspective.

They are going to extend Attribute’s analytics to also include customer cost attribution by the cost per message and utilize the platform as a pivotal decision-making tool for architectural improvements that would have a business impact.

“Their tagless technology made it extremely easy to overcome the data blindspots we had, and we can clearly discover and contextualize how each service or customer utilizes cloud resources, helping us make informed decisions rapidly.”

Ziv Sivan VP of Engineering

Reveal the hidden profit in your cloud spend

Get a demo