As the dominant alternative to Amazon S3, MinIO can be found on-prem, in the cloud or in the colo. Responsible for multiple exabytes of data, MinIO has developed an AI layer called Eureka. With Eureka an organization can quickly start using the latest AI/ML technologies (Large Language Models, Retrieval Augmented Generation (RAG), and vector databases) to apply generative AI to their unstructured data. Given the sensitivity of corporate data, the Eureka approach is optimal for enterprises looking to deliver state of the art capabilities on private and/or sensitive data. 

This talk will discuss the philosophy behind Eureka, the architecture required to operate it at scale and a full demonstration of its capabilities.