OpenAI Goes All-In on RAG with Chat History Search Feature

OpenAI has announced the rollout of a new feature that allows users to search through their chat history on ChatGPT web. This enables users to quickly and easily access previous conversations, making it simple to reference past discussions or pick up where they left off. This feature was developed by Rockset.

Moreover, some users on X  recently reported that ChatGPT is working on integrations with Slack and Google Workspace for enterprise customers, enabling them to add external information to ChatGPT.

OpenAI recently acquired Rockset, a data analytics company co-founded in 2016 by former Meta engineers Dhruba Borthakur and Venkat Venkataramani, for $105 million (INR 905 crore). This acquisition aims to leverage Rockset’s advanced analytics to enhance OpenAI’s retrieval infrastructure.

“We just started rolling out the ability to search for all your past chat history in ChatGPT. This requires real-time indexing and hybrid searches to find the most relevant matches, and what can be a better database to power this than Rockset?” said former Rockset chief Borthakur. 

Rockset is collaborating with OpenAI to tackle the complex database challenges that AI applications face at scale. “We will help OpenAI solve the complex database challenges that AI applications face at a massive scale,” the company said.

Rockset enhances the capabilities of large language models (LLMs) through its retrieval augmented generation (RAG) feature. This integration will be valuable for OpenAI’s enterprise customers who want to utilise models with their proprietary data and reduce hallucination risks. 

By incorporating their data, customers can improve LLMs to deliver more contextual and accurate results, expanding the applications of LLMs in content generation and information retrieval.

Rockset employs a converged indexing approach that combines row, columnar, and search indexes, enabling fast searches across multiple data types and formats. Currently, no other database company, such as MongoDB, Elasticsearch, or Amazon Redshift, offers this specific service.

While Elasticsearch excels in keyword search and alternatives like Weaviate and Pinecone specialise in vector search, Rockset merges these capabilities to provide precise keyword matches alongside semantically rich search results. 

Its hybrid vector search combines traditional keyword search with vector search, allowing for more relevant and context-aware results. This is particularly beneficial for OpenAI, which manages vast volumes of unstructured data, including text, images, and audio.

Unlike some solutions requiring complex infrastructure management, Rockset offers a fully managed, serverless architecture that reduces operational overhead and simplifies scalability. This approach eliminates the need for manual infrastructure management, allowing OpenAI to scale its operations seamlessly and efficiently.