OpenAI has acquired Rockset, a vendor specializing in real-time analytics databases, to enhance its retrieval capabilities. The financial terms of the deal were not disclosed, but OpenAI confirmed that Rockset's indexing and querying capabilities will be integrated into its retrieval infrastructure. Rockset's entire team will join OpenAI as part of the acquisition.
This acquisition is OpenAI's second major public acquisition, following its purchase of Global Illumination, Inc. last year. The move comes amid growing competition in the generative AI space, highlighted by Anthropic's recent release of Claude 3.5 Sonnet, which outperforms OpenAI's GPT-4o on benchmarks. Additionally, OpenAI co-founder Ilya Sutskever has launched a new AI startup, Safe SuperIntelligence.
Rockset's Capabilities
Founded in 2016, Rockset provides a cloud-based real-time analytics database that allows developers to build data-intensive applications. Its product continuously ingests and indexes data from sources like Kafka, MongoDB, DynamoDB, and S3, enabling real-time querying without a predefined schema. Rockset uses the open-source RocksDB persistent key-value store as a foundation, acting as an external secondary index for OLTP databases, data lakes, and streaming platforms. This accelerates real-time analytic queries and provides performance isolation for primary transactional systems.
Integration with OpenAI
The integration of Rockset’s technology into OpenAI aims to enhance the retrieval stack of OpenAI products, enabling them to answer customer questions with the freshest and most relevant information more quickly. This will likely improve the performance of enterprise GPT models in addressing user queries with accurate data.
It remains unclear how Rockset's business with its existing customers, including notable players like Klarna, Meta, Whatnot, and Windward, will be affected. VentureBeat's questions on this matter were unanswered at the time of writing.