Gemini has new features for developers
At the event, the company also made a number of announcements around AI, including Gemini Code Assist, a platform for developers that uses the latest Gemini models to help them build applications with greater speed and quality in popular code editors like VS Code and JetBrains, in addition to supporting private code bases, whether Gitlab, Github, Bitbucket or even through multiple repositories.
Additionally, Gemini Cloud Assist helps cloud teams design, operate, and optimize the lifecycle of their applications, using Gemini's personalized, contextual AI guidance that understands their Google Cloud resources to help them build new designs, deploy workloads, manage applications, troubleshoot, and optimize performance as well as costs.
Among some of the capabilities of this new assistant is goal-oriented design, that is, by describing the desired design result you can generate architectural configurations adapted to your needs. He even explains the reasoning behind his suggestions.
It also features Guided Operations and Troubleshooting that diagnoses, identifies root causes, and recommends solutions, helping you quickly resolve incidents during development, as well as identify areas for improvement and suggest how to achieve goals.
To directly access Gemini Cloud Assist, you only need a chat interface in the Google Cloud console, which is directly integrated into the interfaces where different products and resources are managed in the cloud; However, Gemini Cloud Assist is in private testing for now.
Among other announcements, Google said Gemini 1.5 Pro is now available in public preview on Vertex AI, the end-to-end platform that works not only to build generative AI applications and agents, but also to deploy and maintain them.
In this sense, one of the main changes is that now Vertex AI users will have the ability to base their product responses with Google Search, that is, they will have access to fresh, high-quality information that improves the accuracy of the model responses.
Google also announced that Gemini 1.5 Pro is now in public testing, which is significant for workflows as it provides a context window of 1 million tokens. This advancement, the company says, enables native reasoning on huge amounts of request-specific data, eliminating the need to tune base models or employ Recovery Augmented Generation (RAG) to base model responses on external data.
