Cloud technology: A synergistic environment for KM and generative AI
A partnership between Sinequa and Google Cloud allows Sinequa’ s Assistant, the company’s retrieval-augmented generation (RAG), to support Google Cloud’s Vertex AI platform, a generative AI (GenAI) implementation. “Combining search with GenAI means accurate, traceable, and up-to-date information,” observed Evernham. “This eliminates hallucinations by connecting GenAI to established corporate knowledge.” GenAI and search are complementary, he pointed out. “Instead of the user having to read and evaluate each of the documents that is retrieved in response to a search, GenAI reads and summarizes them. Search retrieves the appropriate documents and provides citations, so the sources are known and traceable,” Evernham explained.
Customers looking for a search solution nearly always have the same reason for seeking an alternative, and it generally amounts to some harsh words for their existing search capabilities. “Sometimes, they can’t find something they know is there because the query was not precise enough, or they have to search in too many places,” he recounted. “Many applications have only a rudimentary search component, and it’s not enough.” The real solution includes having the proper connectors to repositories, honoring permissions for secure information access, enriching the metadata,
classifying the content, and utilizing large language models (LLMs) to understand the semantics.
A cloud search solution should also include the ability to incorporate internet sites in the search. “In addition to searching data in the user’s cloud,
Sinequa can also index content in sources such as PubMed to identify, for example, all the documents related to oncology,” Evernham added. “This allows users to reach whatever information they need to get the answers they need.”
AI is at a pivotal moment. The cloud provides the best environment for GenAI to thrive, given the resource demands— terabytes of storage and the large number of parameters required for learning models to get accurate results, for example. According to IDC, global AI spending on the public cloud will exceed $228 billion by next year. This synergy can only be expected to increase as more applications emerge.