Google's Vertex AI Is Targeting Enterprise Users with New Features

With the help of Google's vast language models and superior machine learning, Vertex AI, an advanced development service offered by Google Cloud, is bringing new features and improvements to increase the accuracy of AI-driven apps. Since Vertex AI's Grounding with Google Search feature went live in May, Google has released more tools that let users improve their AI results by utilising certain third-party datasets.

These updates are part of Google’s ongoing efforts to encourage the adoption of its “enterprise-ready” generative AI solutions by reducing the frequency of misleading or inaccurate outputs, which we have seen in the part during the AI-powered search disaster.

In addition to these third-party datasets, Google is launching a “high-fidelity mode” for Vertex AI. This mode allows organizations to source information directly from their own company datasets rather than relying solely on Gemini’s comprehensive knowledge base. Powered by a specialized version of Gemini 1.5 Flash, high-fidelity mode is currently previewed through Vertex AI’s Experiments tool.

Another significant update is the expansion of Vector Search, which enables users to find images by referencing similar graphics. This feature now supports hybrid searches, combining vector-based and text-based keyword searches to improve search accuracy. This update is available in public preview.

Furthermore, the dynamic retrieval feature within Grounding with Google Search will soon be able to automatically determine whether information should be sourced from Gemini’s established datasets or from Google Search, particularly for prompts requiring the latest/updated information.

These advancements come at a crucial time as Google seeks to improve the reputation of its AI-based search features. The AI Overviews feature, which utilizes data from platforms like Reddit, previously faced criticism for generating unusual recommendations. With these new capabilities, Google aims to provide more accurate and reliable AI-driven solutions, reinforcing its commitment to high-quality information retrieval and application development. The enhancements to Vertex AI are expected to significantly strengthen the platform’s pull to enterprise users, offering more precise control over information sources and improving overall AI output reliability.