Vectara provides Neural Search as a Service

Built by a team of ex-Google technologists and AI scientists, Vectara operationalizes large language models to deliver intent-based search results that are claimed to radically outperform keyword-based systems

Available in two editions; Growth provides a generous free tier, and Scale includes a complete set of enterprise features. Vectara makes neural network technology immediately accessible to developers, even without any machine learning expertise.

The company has also announced $US20 million in seed funding. Funding partners include tech and search pioneers from companies, including Slack and Google, alongside leading technology venture funds.

Vectara’s platform aims to become the leading engine for unlocking insights and value from the volumes of untapped data held by most modern organizations. To achieve this vision, Vectara developed zero-shot AI models that leverage the power of deep learning neural networks to analyze unstructured data without requiring additional training.

The platform provides API access to the most advanced NLP and AI capabilities in minutes, empowering developers to quickly integrate the platform into their stacks and augment site and app search, product discovery, customer support, user navigation, and more.

Vectara is launching with search because it is the dominant way people navigate applications and sites to access knowledge today. However, organizations often struggle to implement search into their applications as it requires a significant time investment to operationalize while leaving relevant information undiscovered.

Vectara replaces these outdated approaches with a composable, ML-based search pipeline that recognizes searchers’ intent and radically increases search relevance. Vectara’s ability to innately understand human language means that the platform sees through incorrect terminology, acronyms, or misspellings without any predefined language configuration or developer tuning.

Because Vectara applies native neural search, it is human-language agnostic; it can even conduct cross-language searches, surfacing information in one language from content written in another without translation.

“We started Vectara because we saw an opportunity for neural networks to revolutionize how people find meaning through data,” said Amr Awadallah, Vectara’s Chief Executive Officer.

“Until recently, though, only the largest organizations with deep ML expertise could access this breakthrough technology. With our API-first neural search, developers and engineers can immediately start applying NLP to the challenges they face every day. I have no doubt that the future of search is unequivocally neural search.”

Vectara’s turnkey neural search platform is available to users everywhere, with zero language configuration or tuning requirements. Customers get an end-to-end search pipeline with market-leading search relevance, available on a tiered pricing model. The free version, Growth, offers users up to 15,000 queries per month, while Scale provides a complete search pipeline capable of running on the largest websites in the world.