Google Cloud AI

Tietotalo builds AI on Google Cloud with Gemini and the Gemini Enterprise Agent Platform, formerly Vertex AI, when your data and tools already live there.

If your data already lives in Google Cloud, moving it somewhere else just to use AI is the wrong first step.

Tietotalo is best known for Microsoft work, and most of our clients run on Azure and Microsoft 365. But plenty of organisations do not. Some run their analytics in BigQuery, their documents in Google Workspace or their apps on Firebase. For them, the most direct route to useful AI is usually the platform their data is already on, not a migration project with AI attached at the end.

Google's enterprise AI stack changed name and shape in April 2026, when Vertex AI became the Gemini Enterprise Agent Platform. It combines Gemini models, a catalogue of third-party models, tools for building and governing agents, and the Gemini Enterprise app where staff use those agents day to day. We build on it when it is the shorter path to something that works, and we say so when another platform would fit better.

How we help

  • Platform fit assessment: Looking at where your data, identities and existing apps live, and deciding whether Google Cloud, Azure or a direct model API gives you the simplest and cheapest route.
  • Gemini based applications: Document understanding, summarisation, classification and chat over your own content, using Gemini models through the Agent Platform.
  • AI over your BigQuery data: Letting staff ask questions of their analytics data in plain language, with the generated queries visible and checkable rather than hidden behind a chat window.
  • Agents with the Agent Development Kit: Building agents that call your systems and complete tasks, deployed and monitored on Google's managed agent runtime.
  • Google Workspace integration: Connecting AI to Gmail, Drive and Docs content for teams that work in Google Workspace rather than Microsoft 365.
  • Multi cloud connections: Linking Google based AI to systems that run elsewhere, such as a booking platform on Azure or a CRM in Dynamics 365.

Technology

  • Gemini Enterprise Agent Platform (formerly Vertex AI)
  • Gemini models and Gemma open models
  • Model Garden, including Anthropic Claude and other third-party models
  • Agent Studio and Agent Development Kit (ADK)
  • Gemini Enterprise app (formerly Agentspace)
  • BigQuery and BigQuery ML
  • Document AI and Cloud Storage
  • Cloud Run and Cloud Functions
  • Firebase
  • Google Workspace: Gmail, Drive, Docs
  • Cloud IAM and VPC Service Controls

Why it matters

We do not believe in forcing every client onto the Microsoft stack just because that is where most of our projects are. The same principle sits behind our work on Firebase and Amazon AWS: the right platform is the one that fits your data, your team and your existing contracts.

What carries across clouds is the part that actually decides whether an AI project succeeds: a clear use case, clean source data, evaluation against real examples, and someone accountable for the result once it is live. That work is the same whether the model runs in Azure or in Google Cloud.

The trap in Google's AI stack is the naming, not the technology. Gemini, the Gemini app, Gemini Enterprise, the Gemini Enterprise Agent Platform and Gemini in Workspace are different products with different licences and different data terms. It is easy to assume that because staff already have Gemini in Workspace, an internal agent built on Google Cloud is covered by the same agreement. Before any build, we map which product handles which data under which terms, so the compliance question is settled before the first prompt is written.

Frequently asked questions

Vertex AI was rebranded as the Gemini Enterprise Agent Platform at Google Cloud Next in April 2026. Existing APIs keep working, but Google now delivers new features under the Agent Platform name.
Because the right platform is the one that fits your data and your team. If everything you have runs on Google Cloud and Workspace, building on Azure would add cost and complexity for no real benefit.
Gemini is a strong default on Google Cloud, particularly for long documents and images. The Agent Platform also offers third-party models, including Claude, so we can test alternatives without leaving your Google environment.
Yes. Models and agents on Google Cloud are reached through APIs and can call APIs themselves, so connecting to Dynamics 365, SharePoint or a booking system is an integration task like any other.
Google Cloud lets you choose regions, including EU regions, for most AI services. Availability differs per model, so we confirm the region and data processing terms for each model before building.