Google has begun letting ordinary people build their own AI helpers inside the tools they already use for work, through its Workspace suite, without writing a line of code. The idea is an old one, the simple promise of automation that says when something happens, do this, now carried further by a system that can read, weigh and respond to what it finds. Its arrival is measured in weeks rather than a single day, and across the wider story of office software it marks a shift from people instructing machines step by step toward describing intentions and letting the machine work out the steps. Whe
Google Launches Workspace Studio, Letting Users Build No-Code AI Agents Powered by Gemini 3
Agents built in minutes, from plain-language prompts, with no code
Mark
So what actually is Workspace Studio?
Mimi
It's a tool inside Google Workspace for building custom AI agents with plain-language prompts, no code. They run on Gemini 3.
Mark
How is that different from ordinary automation?
Mimi
It still uses an if-this-then-that structure, but Google says Gemini lets agents read content, detect sentiment and adapt to new information instead of following fixed rules.
Luke
Worth flagging that last part is Google's claim. The reporting doesn't show examples of agents adapting well or badly.
Mark
What are the building blocks?
Mimi
Starters, steps and variables. A starter is the trigger, like a new email from a particular person. Steps are the actions, like drafting a reply. Variables hold things like names, dates or invoice numbers so they can be reused down the workflow.
Mark
And who gets it?
Mimi
Business, Enterprise and Education plans, plus Google AI Pro for Education and AI Ultra for Business. It's rolling out over the coming weeks.
Luke
So not everyone has it today. The article gives no timeline beyond "coming weeks" and no usage numbers.
Mark
Is there anything on reliability or safeguards?
Luke
Nothing. How accurate the drafted replies are, or what oversight users get, is unknown from this piece.
Mimi
What we do know is where it lives: a shortcut beside the Gemini icon in each major app, and a dashboard at studio.workspace.google.com.
Der Puls
- Google has started rolling out Workspace Studio, letting anyone build a Gemini-powered AI agent in minutes using plain-language instructions.
- The tool promises to move past rigid 'if this, then that' rules, with agents that interpret content, judge sentiment and adapt to new information.
- Agents chain together triggers, steps and variables, so a single workflow can pull names, dates or invoice numbers and act on them across emails and documents.
- The rollout is staggered over the coming weeks across Business, Enterprise, Education and Google AI Pro and Ultra tiers, leaving adoption and reliability still unproven.
Google has begun letting ordinary people build their own AI helpers inside the tools they already use for work, through its Workspace suite, without writing a line of code. The idea is an old one, the simple promise of automation that says when something happens, do this, now carried further by a system that can read, weigh and respond to what it finds. Its arrival is measured in weeks rather than a single day, and across the wider story of office software it marks a shift from people instructing machines step by step toward describing intentions and letting the machine work out the steps. Whether these agents prove dependable in the messy reality of real inboxes and documents is a question the coming months will answer.
Google has started rolling out Workspace Studio, a tool that lets people build their own AI agents inside Google Workspace without writing code. The agents run on Gemini 3, and Google says a working one can be assembled in a few minutes from ordinary written instructions.
The product has been a long time coming. It was first previewed at I/O 2024 and announced again in April, and now it reaches a wider group of Workspace customers, with the rollout unfolding over several weeks rather than all at once.
Each agent is built from three parts. Starters define what triggers it, such as a set time or an email from a particular contact. Steps describe what happens next, like drafting a reply or updating a document. Variables carry extracted details, names, dates or invoice numbers, through the workflow. Google says the agents go beyond fixed rules, interpreting content and adapting to new information, and users can share them as easily as files on Google Drive.
A shortcut sits beside the Gemini icon in every major Workspace app, and a standalone dashboard is available for those who want more room to work.
What remains to be seen is how widely people adopt it and how reliably the agents behave once they are handling real email and documents. For now, these capabilities are Google's own account, and the rollout has only just begun.
Google has begun rolling out Workspace Studio, a tool that lets people build their own AI agents inside Google Workspace without writing any code. The agents run on Gemini 3, and the company says they can be put together in a few minutes using ordinary written instructions.
The product has been a while in coming. It was first previewed at I/O 2024 and announced again in April. Now it is reaching a wider set of Workspace customers, with the rollout set to unfold over the coming weeks rather than all at once.
Workspace Studio follows the familiar "if this, then that" model of automation, but Google says it adds Gemini's reasoning and multimodal understanding on top. In practice, the agents can process emails, analyze content, extract information, prioritize important tasks and send notifications. Among the uses Google points to are sending emails and drafting replies.
Each agent is built from three components. Starters define what sets the automation off, such as a certain time or a new email from a specific contact. Steps spell out what the agent should then do, for example drafting replies, organizing information or updating documents. Variables act as placeholders, letting an agent carry extracted data through the workflow: names, dates, invoice numbers, or responses that Gemini has generated.
Google argues this goes beyond rule-based automation. According to the company, agents can interpret content, detect sentiment and adapt to new information, rather than simply matching fixed conditions. Users can also share the agents they build in the same way they share files on Google Drive.
Access is designed to be close at hand. Google has placed a shortcut in every major Workspace app, next to the Gemini icon. The interface includes Discover, My Agents and Activity tabs, and a standalone dashboard lives at studio.workspace.google.com.
The feature is slated for Business Starter, Standard and Plus; Enterprise Starter, Standard and Plus; Education Fundamentals, Standard and Plus; Google AI Pro for Education; and the Google AI Ultra for Business tier.
What remains to be seen is how widely people adopt it and how reliably the agents behave once they are handling real email and documents. For now, the capabilities described are Google's own account of the product, and the rollout is only beginning.
Bemerkenswerte Zitate
Google says the system goes beyond rule-based automation by letting agents interpret content, detect sentiment and adapt to new information.— Google, as reported by NewsX
Möchten Sie die ganze Geschichte?
Das Original lesen bei NewsX ↗
Technology