Google Cloud

Build and deploy AI agents

Gemini Enterprise Agent Platform allows developers to build, scale, govern, and optimize enterprise-grade AI agents. This feature provides a comprehensive environment for transforming applications and workflows into powerful agentic systems, enabling rapid development and deployment.

Train custom Models

This feature provides a complete overview of the custom training workflow, allowing users to control the training process using their preferred ML framework. It includes steps from data preparation to model predictions, ensuring flexibility and customization.

Deploy a model for production use

Users can register their models to the Model Registry and utilize the Agent Platform prediction service for both batch and online predictions. This feature simplifies the deployment process, making it easier to integrate models into real-world applications.

Extract, summarize, and classify data

This feature enables users to create text prompts for various tasks such as classification, summarization, and data extraction. It provides flexibility in designing prompts, allowing for tailored interactions with the generative AI capabilities of the platform.

200+ Google and third-party AI models and tools

Users can choose from a wide array of multimodal models, including Google's latest offerings and third-party models. This feature allows for customization and tuning of models to fit specific use cases, enhancing the versatility of the platform.

MLOps for predictive and generative AI

The platform provides specialized MLOps tools for data scientists and ML engineers to automate and manage ML projects. This includes features for model evaluation, orchestration of workflows, and monitoring models for performance.

Open and integrated AI platform

Data scientists can leverage integrated tools for training, tuning, and deploying ML models. The platform supports seamless integration with BigQuery and other Google Cloud services, facilitating a unified workflow across data and AI tasks.

Agent Search for healthcare

This feature provides a medically-tuned search experience that enhances the retrieval of relevant information from structured and unstructured patient records. It integrates with Healthcare Data Engine to streamline the search process for healthcare professionals.

DIY with Vector Search and embeddings

Agent Search allows developers to create custom search engines using embeddings for semantic matching. This feature supports building recommendation engines and other applications that require advanced search capabilities.