SANTA CLARA, Calif., May 30 (Korea Bizwrie) – Vectara, the Generative AI (GenAI) conversational search platform, has established itself as a leading player in GenAI with the release of an all-new “Grounded Generation” capability that all but eliminates hallucinations. With this release, Vectara empowers developers to rapidly and easily build conversational AI products with world-class retrieval, summarization, and data privacy directly on top of the data that matters to their business. Vectara raised $28.5 million in Seed funding to accelerate its growth trajectory, led by Race Capital, and established a new Strategic Board of Advisors, made up of principal minds in the space, including Matei Zaharia, the Cofounder and Chief Technologist at Databricks.
Hybrid Search & Beyond: Introducing “Grounded Generation”
Vectara’s new Grounded Generation, or Retrieval Augmented Generation, raises the bar as it allows users to ask questions about their data and returns a natural language summary based on relevant facts retrieved with citations to the company’s source data set. These citations support fact-checking and substantially reduce hallucinations, providing business users with the peace of mind that Vectara will only base its generative responses on the factual information – documents and other data – that they’ve provided to Vectara.
In addition to countering hallucination, Vectara is also improving search efficacy with its enhanced Hybrid Search. Vectara’s Hybrid Search uses a combination of semantic search, Boolean, and exact keyword matching approaches to provide users with the most relevant answers regardless of a query’s length, level of ambiguity, provided context, or even language used.
Unlike many API providers, Vectara does not train its models on its customers’ data, such as indexed data or queries. By also providing client-configurable data retention, Vectara enables organizations to discard the original documents and text after they have been indexed so that no residual data from the company remains in the index. Vectara’s instant indexing ensures data is updated in near real-time.
“These new features and capabilities make Vectara’s neural retrieval platform among the best in the world,” said CEO and Cofounder Amr Awadallah, “The breakthroughs our team has accomplished over the last eight months are changing the face of AI and how companies can safely use it to expand and improve their value propositions. The new Summarization feature provides organizations with trustworthy ChatGPT-like generated answers, stacked on top of our Hybrid Search enhancements for superior search efficacy. This release makes Vectara one of the most advanced companies in the world in information retrieval and Grounded Generation.”
Funding for Mission-Driven Innovation
Built by a team of ex-Google AI Research technologists and AI scientists, Vectara empowers developers to embed the most advanced Natural Language Understanding models for question answering in minutes. Vectara is making sophisticated AI more accessible for developers and engineers who aren’t trained to build AI from scratch or prepared to invest in the infrastructure and model maintenance overhead. In closing its Seed Funding Round at $28.5 million, Vectara is poised to expand its strategic differentiators – superior information retrieval and ease of use – to further democratize this impactful technology.
“At Race Capital, we are thrilled to lead Vectara’s latest seed round and are committed to working closely with Amr, Amin, Tallat and the amazing team they assembled to further push the boundaries of what’s possible with grounded generative AI,” said Alfred Chuang, Founder and General Partner at Race Capital. “AI can potentially generate trillions of dollars in economic value, while the destructive influence of hallucination holds the power to undo it all. Accuracy and data security are crucial for many industries including financial, healthcare and education. Vectara will be their go-to solution.”
Emad Mostaque, CEO and Founder of Stability AI, is also supporting Vectara’s innovations. “I am delighted to be personally included as one of the angel investors for Vectara. Their ML researchers truly understand how to leverage large language models, as many of them worked on these technologies at Google AI research. I look forward to witnessing Vectara transform the conversational knowledge discovery experience by leveraging generative AI to its fullest potential.”
New Strategic Board of Advisors
Vectara was founded on the prediction that within five years, every company will move beyond search to LLM-powered conversations with owned data assets. To dominate the $51.8 billion market for LLM applications, Vectara has established a new Strategic Board of Advisors. This Board provides Vectara with counsel and expertise from tested veterans in the space, from companies and institutions including Databricks, Cloudera, Stanford University, and Northeastern University.
“Vectara’s new funding round will be instrumental in developing and expanding access to the platform’s breakthrough hybrid search and grounded generative AI features,” said Databricks Chief Technologist and Cofounder Matei Zaharia. “Databricks Ventures believes wholeheartedly in Vectara’s vision for the future of data intelligence enhanced by LLMs. As a new Vectara Board of Advisors member, I look forward to supporting their product roadmap and global market strategy as the company’s growth accelerates.”
Vectara is a conversational GenAI search and discovery platform, which allows businesses to have intelligent conversations over their own data (think ChatGPT but for your data). Developer-first, the platform provides a simple API and gives developers access to cutting-edge NLU (Natural Language Understanding) technology with industry-leading relevance. With Vectara’s Grounded Generation, businesses can quickly, safely, and affordably integrate best-in-class search and question answering into their application, knowledge base, site, chatbot, or support helpdesk. Visit Vectara.com for more information.
Source: Vectara via GLOBE NEWSWIRE