Vectara

Vectara is a generative AI platform provider founded in 2022 that offers Retrieval Augmented Generation-as-a-Service (RAGaaS) to enterprises. The company's core product is a cloud-based conversational generative large language model "search-as-a-service" that allows businesses to hold intelligent conversations with their own data, such as documents, knowledge bases and code. Vectara's proprietary AI technology works on businesses' own data, providing an application programming interface that gives developers quick access to embed its capabilities into apps, websites, chatbots and help desks.

A key feature of Vectara's platform is "Grounded Generation", which augments queries with facts and data related to a company's own information to reduce AI "hallucinations" - incorrect or random responses. The platform combines multiple search methods including semantic search, Boolean and exact keyword matching to provide highly relevant responses, even with misspellings, alternate meanings and different languages. Vectara can extract information from numerous document types and code formats. In May 2023, Vectara introduced a factual consistency score metric to assess the accuracy of AI-generated responses. The company's Hughes Hallucination Evaluation Model (HHEM) compares original source documents with generated summaries to assign a factual consistency probability score.

In February 2024, Vectara launched a Chat module to empower companies to build advanced chatbot systems. The module offers features like progressive conversations with full chat history context, a white-label UI chat widget framework, and user behavior trend analysis. In July 2024, Vectara unveiled Mockingbird, a fine-tuned generative Large Language Model specifically designed for RAG applications to reduce hallucinations and improve structured output.