Case Study
CertiK, a leader in blockchain security technology, was experiencing more than a 9X growth of Snowflake costs in only six months, vastly outpacing their planned budget. Through a two-week low friction trial, Bluesky helped CertiK save hundreds of thousands of dollars by pinpointing the specific queries that were responsible for the cost overruns, generating tuning insights for these queries that led to over 300X cost reduction and 500X speed improvement, and providing continuous monitoring for emerging expensive queries to help CertiK maintain data cloud efficiency.
CertiK has multiple Snowflake warehouses with each dedicated to a different use case such as data ingestion, data aggregation, serving, and data collection from observability tools. They first started using Snowflake in the middle of 2021 and costs were negligible. Soon data engineers found Snowflake extremely easy to use and started to add more and more new use cases including innovative yet expensive use cases such as analysis of real-time on-chain data. Consequently, costs went up exponentially. In nine months, costs reached an annual run-rate that exceeded a quarter of a million dollars — which is when CertiK decided to seek help from Bluesky for workload efficiency optimization and cost reduction.
Bluesky delivers an intelligent workload optimization and cost governance solution for modern data cloud environments, helping Snowflake users continuously improve workload performance, optimize costs for superior financial management and accelerate data-driven innovation.
CertiK went through a trial by granting Bluesky the READ ONLY access to metrics data under the Snowflake database schema and query history. The entire onboarding process took less than ten minutes. Bluesky does not require access to any user data in order to:
Bluesky then processes metrics and query history to identify cost reduction opportunities.
“While we enjoyed the raw power and ease-of-use with Snowflake, we quickly ran up our expenses. Fortunately, through an easy and collaborative production trial and adoption, Bluesky was able to help identify and optimize costly queries, thus reducing our costs meaningfully." Justin Zhang, Sr. Director of Engineering and Head of Data, CertiK
Within the first two weeks, Bluesky identified cost reduction opportunities that equal to ~20% of CertiK’s overall Snowflake spend. For example, It identified a QA workload that would have cost more than $300K annually and shut it down; it found certain expressive SQL queries without the right filtering conditions and generated alternative query formulations to reduce the costs by more than 300X and improve query speed by 500X; it also computed optimal warehouse settings that led to further cost efficiency improvement.
Since completing this initial significant milestone, CertiK has been monitoring the evolving Snowflake workloads and identifying emerging expensive and inefficient query patterns using Bluesky’s alerting feature. CertiK’s data engineering team now is able to stay informed and react to changing Snowflake workloads in a timely manner, and avoid queries similar to the $300k one from happening again. Additionally, the team can focus on strategic initiatives that further maximize the ROI of Snowflake.
CertiK plans to continue monitoring and optimizing its Snowflake data warehouse infrastructure using Bluesky. CertiK and Bluesky will also continue their strong partnership to build out Bluesky’s next-generation Big Data efficiency platform - Bluesky Intelligent Compute - which will help autotune queries and intelligently route workloads to the right warehouse for maximum efficiency.