Proven with Industry Leaders
Which Google AI tool does which job
Five names cover most of what enterprises ask about. Each serves a different user, and each needs its own controls.
| Tool | What it is for | Who uses it |
|---|---|---|
| Gemini Enterprise | The employee app for using and discovering agents | Business users |
| Gemini Enterprise Agent Platform (formerly Vertex AI) | Building, scaling, governing and optimizing agents and models, with access to Google and third-party models | AI and platform teams |
| Agent Development Kit (ADK) | The code-first way to build agents | Developers |
| BigQuery data agents | Conversational Analytics, the Data Engineering Agent and the Data Science Agent, working inside BigQuery | Data teams |
| Data Agent Kit (preview) | Agent skills for data work inside developer tools such as Claude Code and VS Code | Data engineers |
Product names and release status as described by Google in its Agentic Data Cloud announcement, checked November 2026.
Govern agents before go-live
Google's platform provides the controls. The work is connecting them to the way your organization already approves and audits change.
Record every agent
Each agent goes in your AI inventory, with its owner, model, data and tools.
AI governance consultingEvaluate before release
Evaluate before release and after every model change. Google's rename of Vertex AI is itself a change your records should note.
Give agents tools through governed APIs
MCP behind an API gateway, on WSO2, Apigee or the gateway you already run, with every call logged.
AI gateway and agent governanceKeep evidence your team can show
Prompts, tool calls, approvals and incidents, kept in a form your risk and audit teams can review.
Decide which Google tools fit, and what to govern first
Bring the use cases you are weighing. Leave with a view on which Google tools fit them and which controls to put in place first.
- What it is: a conversation, not a sales call.
- Timing: we reply within one business day to set a time.
- Prefer evidence first? The other tab maps the data behind one AI use case, free for a closely defined use case.
Google's AI tools, answered plainly.
Is Vertex AI gone?
Google announced Gemini Enterprise Agent Platform as the evolution of Vertex AI in April 2026. Model building, tuning and serving continue under the new platform, and new agent features ship there. Check Google's notes for your workloads.
What is the difference between Gemini Enterprise and Gemini Enterprise Agent Platform?
Gemini Enterprise is the app employees use. Agent Platform is where technical teams build, run and govern the agents that appear in it.
Can we use models other than Gemini?
Yes. Agent Platform includes Model Garden, which offers Google and third-party models.
Do our agents need MCP?
Agents need a governed way to call your systems. MCP is a widely adopted protocol for that. We put it behind an API gateway so access, limits and logs follow your existing rules.
How do you govern agents on Google Cloud?
With an inventory, evaluation before and after changes, tool access through governed APIs, and a running record of what each agent did. Google's platform provides controls. We connect them to your model risk and audit process.