• Google introduces AI-focused features in app development and infrastructure.
• Event features Ironwood, a TPU built for AI inference.

Google has introduced several updates aimed at simplifying application development and expanding network capabilities – many of which are now supported by AI.
At the company’s Google Cloud Next 2025, it showed its intentions to try and shift developers’ experiences away from managing infrastructure towards working directly with applications.
Traditionally, cloud teams handle tasks like provisioning servers and managing storage before building software. Google’s new tools aim to streamline such processes. The Application Design Centre, now in public preview, gives developers a visual space in which to create and configure application templates. It includes features like drag-and-drop components, in-line infrastructure-as-code, and one-click deployments.
Projects so created connect directly to App Hub, a platform that helps monitor application health and track deployments across cloud environments.

Google also previewed Cloud Hub, an interface that consolidates application insights – from resource use and system health to billing and quotas. It’s designed to simplify how developers and operators manage workloads in different service regions, and is now accessible in the Google Cloud console.

AI throughout the development lifecycle

AI to become more embedded in Google’s cloud tools. Gemini Code Assist now integrates with popular IDEs like Android Studio, VSCode, Firebase Studio, and JetBrains.

Developers can use Cloud Assist to automate tasks like code generation, debugging, documentation, and translating between programming languages. A kanban board shows active tasks, allowing teams to track AI-generated work in real-time.

On the mobile development front, Firebase Studio includes AI agents that turn user prompts into working prototypes. The agents can generate interface elements, write backend logic, and create and run tests – automating much of the early app development process.

Gemini Cloud Assist will provide AI support for infrastructure management. It’s designed to guide users through system planning, troubleshooting, and cost optimisation. A new feature called Investigations, now in private preview, examines logs, configuration changes, and telemetry data to help pinpoint problems. Teams can hand its findings off to Google Support for additional help if needed.

Cost management also got some attention. Cost Explorer helps teams connect resource use to spending. The dashboard reviews 30 days of compute and memory use, helping identify underused services and offering cost-saving recommendations.

Expanding global infrastructure with Cloud WAN

For organisations with global operations, Google is extending its network capabilities with Cloud WAN. The managed service is built on Google’s backbone network and is intended to replace traditional wide area networking tools like MPLS.

A key component is Cross-Site Interconnect (CSI), which offers dedicated Layer 2 connections between data centres using 10G or 100G links. Now in preview, CSI is designed for enterprises, telecoms, and government agencies that need reliable, high-performance connections across regions. Google’s Premium Tier routes traffic through the nearest point of presence to minimise latency.

The service also integrates with third-party security and connectivity vendors like Cisco, Fortinet, and Palo Alto Networks, allowing organisations to manage networking and security together. Pricing options include both use-based and fixed models, and Google says the offering could cut total cost of ownership by up to 40% compared to managing WANs in-house.

Introducing new hardware at the Cloud Next 25

Google also introduced Ironwood at the event, a seventh-generation Tensor Processing Unit (TPU), designed for high-performance AI inference.

Ironwood is designed to handle demanding workloads like large language models and real-time applications, resulting in better speed, scalability, and energy efficiency. It integrates with Google’s Pathways system and will support tools like Gemini 2.5 and AlphaFold.