Pinecone

Pinecone offers a serverless vector database that can transform and index a large number of high dimensional vectors to search across multiple applications and answer queries as the nearest neighbor search (NNS). The company’s technology leverages AI-generated representations of unstructured data that encapsulate the meaning of the original content in a machine-readable format to enable its users to store and search for data using a number of methods such as keywords, semantic text, multimodal, and image retrieval.

In September 2021, the company launched “Pinecone 2.0,” which included features such as hybrid memory, disk storage, granular search control, low latencies, and what it claims to be up to a 10x infrastructure cost reduction. Users can store metadata such as a topic, an author, and/or a category with each item and filter vector searches by this metadata in a single stage to access a higher degree of control over search results and eliminate the need for slow filtering.

In May 2024, Pinecone publicly launched Pinecone Serverless, its new serverless vector database product that allows enterprises to build more complex AI applications and reduce costs, following a successful beta launch with over 20,000 organizations using the product. Its main features include separating reads, writes, and storage, enabling vector clustering atop blob storage to support large volumes of data, introducing new indexing and retrieval algorithms for fast vector search across this blob storage, and a multi-tenant compute layer.

Key customers and partnerships

Key customers include Microsoft, Notion, ClickUp, Gong, HubSpot, Klarna, and Disco.