The catalyst for ChaosSearch was simple, but profound. Founder Thomas Hazel had spent 25+ years designing, building, and inventing new technology and science in the areas of large scale distributed systems and databases. Through this journey, he saw a need to reinvent how information is both represented and analyzed. Why? The tsunami of data being generated is simply outpacing Moore’s Law and the existing computer science.
Thomas’ conclusion: The existing data structures and algorithms used to store and access information would always result in costly and complex systems, no matter how the components were combined. The eureka moment was when Thomas mathematically worked out how to singularly and fully index data, for large scale and high performance multi-model analytics, without exploding the size of the
data representation or forcing complex sharding.
The insights culminated in a new representation (think data format) that provides extreme compression, but does not require massive compute. A representation that supports multiple access models such as text search and relational analytics. A representation that is uniquely designed to take advantage of pure cloud object storage. A representation that naturally handles complex and sparse schemas. A representation that is 100% stream-based requiring no sharding, for a new kind of distributed architecture.
With this breakthrough and a seed of an idea, ChaosSearch was born in 2017. Today, its hundreds of employees rally around a shared mission to build the next generation data platform for massive scale analytics that’s extremely simple, cost-efficient, and fast.