A guide to workplace

transformation with the

Gemini Enterprise app
Table of
Contents
01 How to unlock real business value with AI 02
Tackle the cost of data silos with agentic AI

02 Transform work with an agentic platform 04

Gemini Enterprise solutions for your organization

Enterprise Search and AI Assistant

Pre-built tools and agents

Browse and create agents

Connect your apps and data

Security, data governance, and policy controls

Example use-cases

Upskilling your workforce with AI

Future-proof AI strategy with partners and open industry standards

AI that delivers at all levels, by Google

03 Getting started 25

1
 Contact sales

01
How to unlock
real business
value with AI
 Contact sales

Tackle the cost of data

silos with agentic AI
Data is trapped across tools and applications. Research shows that developers using AI
That makes finding and sharing information and assistants to reduce context switching saw their
insights harder than it should be. It also gets in productivity increase by an average of 35%.

the way of decisions being made and actions

being taken. Not to mention that switching apps The friction isn’t in the people or the intelligence
to find all this siloed information costs each of our tools, but in the boundaries between
employee in your organization hours of focus them. People can get better at switching context,
every week. All of that adds up to real time 
 but there’s only so much efficiency in that. What
and money.

your organization really needs is something that

simplifies getting the most from your data.

This constant jumping around is a drain. It breaks

flow, increases errors, and can lead to burnout.

35%
average increase 

in productivity of
developers using AI
assistants to reduce
context switching

3
 Contact sales

02
Transform 

work with an
agentic platform

1
 Contact sales

Solving the data silo challenge

AI can help, but to get the most out of AI, you need a platform
that brings data together from wherever it lives. Gemini
Enterprise app is designed to sit above—and connect with—your
company’s dispersed data to help you find information, and
automate work with agents across the silos.

It’s built to solve the “boundary problem” by allowing your teams

to discover, create, share, and run AI agents in a single, secure
environment. Let’s have a closer look at how the Gemini
Enterprise app is built and how it works.

Gemini Enterprise app

solutions for your organization
Enterprise search and AI assistance
Gemini Enterprise app provides Google-quality search for
your organization’s internal data with the reasoning
capabilities of Gemini models. It lets you retrieve facts
and synthesize answers from across your organization,
rather than simply listing links. It has pre-built connectors
for Google Workspace, Microsoft 365, and applications
like Salesforce, ServiceNow, Jira, and Confluence.

By acting as a secure “connective layer” over existing

platforms and databases, it allows people to query emails,
contracts, and technical documentation simultaneously
from a single interface.

When you ask a question, the assistant provides a

summary based on the information it can find. You can
then ask follow-up questions to refine the answer and get
details based on the data that Gemini searches across.

You can also provide your own context by uploading files

for the assistant to analyze, or by mentioning specific
people and documents in your prompt.

User-directed saved memories allow people to explicitly

instruct the AI to save or discard specific facts. This
enables individuals to override inferred preferences 

and tailor the AI to their exact working style.

5
 Contact sales

Pre-built tools and agents

Gemini Enterprise app comes with specialized, pre-built tools
and agents ready to deliver value for your teams on day one.

Deep Research

Deep Research is designed to automate complex information gathering and synthesis.

Working as a research assistant, it autonomously executes a multi-step research plan
when prompted. It searches across external web sources and internal company data
to read dozens of documents. After filtering out irrelevant noise, it synthesizes the key
points into a comprehensive, fully cited report. You can find Deep Research in the
Agents tab of Gemini Enterprise app. To get started, simply click on it and enter a
natural language prompt defining your research topic.

Getting started Value

Click on it and enter your prompt using natural Instead of spending weeks compiling a
language, defining your research topic. competitor analysis, a strategist prompts,
“Analyze the pricing models of our top 5
competitors,” and gets a cited report in minutes.

NotebookLM

NotebookLM is a specialized tool to which you upload your preferred materials

on a topic, like reports, research papers, YouTube videos, and more. It instantly
becomes an expert on your information. You can then ask questions, get audio or
video summaries, create slide decks, FAQs, and more. You can find NotebookLM
in the Agents tab of Gemini Enterprise app. To get started, simply select it, load
up your sources, and start interacting with the agent using natural language or by
creating different materials (audio, video, etc.) with a few clicks.

Getting started Value

Select it, load up your sources, and start A busy executive loads a notebook with the
interacting with the agent using natural most recent market reports and generates 

language or by creating different materials an audio overview. They can then catch up 

(audio, video, etc.) with a few clicks. on rising trends while commuting, saving 

time and retaining core information.

6
 Contact sales

Idea Generation

Idea Generation is a pre-built agent designed to facilitate brainstorming and

problem-solving. It generates a list of concepts and rigorously ranks them against
each other to find the best options. This allows teams to quickly filter through
noise and identify the most viable solutions for business challenges. You can find
Idea Generation in the Agents tab of Gemini Enterprise app. To get started,
simply select the agent and start brainstorming with a prompt.

Getting started Value

Select the agent and start brainstorming 
 A brand marketer can use this on the spot to
with a prompt. brainstorm fresh campaign angles for a new
product, saving time and energy for
themselves and the team.

Data Insights

Data Insights is a specialized, pre-built agent designed to facilitate data analytics.

It allows people to “talk” to their data using natural language, without the need for
spreadsheet formulas or SQL queries. It can execute data analysis tasks that you
describe using your raw data files, such as CSVs, Google Sheets, or Excel files. 

It can identify patterns, generate charts, and transform datasets into actionable
intelligence. You can find Data Insights in the Agents tab of Gemini Enterprise app.
To get started, simply select the agent, upload your dataset, and use plain
language prompts to visualize trends and uncover insights.

Getting started Value

Select the agent, upload your dataset, and use A finance manager asks, “Show me the 

natural language prompts to visualize trends trend of T&E spend by department for Q3,” 

and uncover insights. and the agent generates a chart and a
summary narrative.

7
 Contact sales

Browse and create agents

with Gemini Enterprise app
Agent Gallery Agent Designer
Agent Gallery is your central hub for discovering, Agent Designer is a no-code platform for
using, and managing agents. To access the gallery, creating, managing, and launching agents
go to Gemini Enterprise app and click Agents in the in the Gemini Enterprise app. With Agent
navigation menu. The gallery displays all available Designer, you can:
agents, organized into these categories:
With Agent Designer, you can:
Create and preview custom agents using 

Made by Google natural language prompts
Ready-to-use and serving a specific Orchestrate complex tasks using multi-step
function, such as Deep Research.
agents (agents with subagents)
Connect your agents to Google and 

From your organization third-party data sources and tools
Agents that your organization has created Schedule agent executions to run tasks 

or shared with you. on a recurring basis
Utilize predictive and deterministic logic 

to ensure reliable agent behavior
Your agents
Incorporate ‘human-in-the-loop’ nodes 

Custom agents that you can create 

using Agent Designer.
to maintain final approval and oversight 

over critical actions

You can use Agent Gallery to:

Select and use any available agent 


to start a conversation and get help
Create a new custom agent using 

Agent Designer
Edit or delete the custom agents 

you have created

8
 Contact sales

Agent Designer canvas

The Agent Designer canvas is an interactive, unified interface for building and
configuring your agent. The canvas is divided into two main interfaces:

Chat interface Designer interface

Provides a conversational interface to build Provides you with granular, low-code control
and refine your agent using natural language over your agent’s configuration and is
prompts. It’s perfect for no-code users who organized into three tabs:
want to use Gemini’s intelligence to quickly
develop and adjust their agent’s details, Flow
instructions, or behaviors. Presents a visual representation of your
agent’s workflow and control logic. Use
the Flow tab to view the overall details
of your main agent, manage the main
agent and all individual subagents, and
to design complex, multi-step agents.

Schedule
Provides configuration options to set
one or more execution schedules for
your agent.

Preview
Provides a live, interactive test
environment for your agent. Use the
Preview tab to test the agent’s
functionality and response as you build

9
 Contact sales

Skills Long-running agents

You don’t need a complex, autonomous agent For complex, multi-step tasks that unfold over
for every task. With Skills, you can create highly hours or days. These agents execute in a
scalable shortcuts for simpler workflows you secure sandbox, can autonomously write code
developed. You can invoke specialized to evaluate their work, and support real-time
instructions and deep expertise in a standard, steering. You can keep a human-in-the-loop
consistent way, without cluttering the model’s approach and guide these long-running agents
core reasoning logic. throughout the task they’re handling.

Value Value
You became really good at quickly updating a Imagine an agent that, over many days,
specific spreadsheet and creating slides monitors inventory, emails approved vendors
weekly to update your team. With Skills, you for quotes, and compiles a comparison table.
can “record” yourself doing it once, and then Once you approve the purchase order, the
invoke the skill weekly so it does it for you. agent submits it, tracks the shipment, follows
up on vendor delays, and updates the ERP
system when the delivery arrives.

Agent Marketplace
The Agent Marketplace allows users to browse and request specialized third-party
agents from partners like Box, Oracle, and ServiceNow, enabling you to deploy
partner-built expertise directly within the Gemini Enterprise app.

10
 Contact sales

Collaboration 

and workflow
These features make collaboration and productivity inside the Gemini Enterprise app
much easier and helpful. From being alerted when an agent completes a task to
being able to collaborate with teammates and agents in real time, the Gemini
Enterprise app helps you get better work done.

Inbox Canvas Projects

A unified command center An interactive editor directly Persistent workspaces that

to monitor all agent activity. within Gemini Enterprise anchor files, conversation
Track progress, receive app that lets you create 
 history, and human-AI
alerts, and provide critical and edit Docs and Slides. collaboration together,
human-in-the-loop input or Collaborate with AI to 
 ensuring all team members
feedback across your refine content and export have access to the same
communication channels. directly to Workspace 
 shared materials and
or Microsoft 365. context.

11
 Contact sales

Connect your apps and 


data to Gemini Enterprise app
Pre-built connectors
An AI model is only as good as the data it can access.
That’s why Gemini Enterprise app connects with the
organizational data of your choice to ensure answers
are relevant and factual. This process is known as
“grounding.” Equally important, you can easily
deselect data sources in case you don’t want agents
to have access.

Gemini Enterprise app includes pre-built connectors

for commonly used applications like Google
Workspace, Microsoft 365, Salesforce, ServiceNow,
Jira, Confluence, Slack, Box, Dropbox, and Zendesk.
This provides you with a multimodal search interface
with permissions-aware access to enterprise
information. More connectors are in the works.

If a user asks, “What is the status of the Project Alpha

contract?”, Gemini will pull answers from actual
sources and cite them, instead of inventing something
that sounds convincing but could be wrong.

You can also build custom connectors, which let you

integrate external data sources that fall outside the
Gemini Enterprise app standard library, making your
organization’s unique data searchable and accessible.

For sources like Google Drive, Gemini uses a zero-

copy approach, meaning it retrieves information
directly from the source at the time of the query
without copying the data into a separate search index.

12
 Contact sales

Bring your own MCP (BYO-MCP)

Admins can register remote Model Context Protocol servers via public URL and
OAuth. This allows teams to select custom databases as secure data stores for their
agents, deepening the grounding of AI responses.

Gemini Enterprise app integration

with Microsoft 365
If your organization runs on Microsoft 365, Gemini Enterprise app integrates directly
into that ecosystem. It includes pre-built connectors for Outlook, SharePoint Online,
OneDrive, and Microsoft Teams. This allows you to overlay Gemini’s powerful AI
models and reasoning onto your existing Microsoft data without complex migrations.

For core Microsoft services like OneDrive and SharePoint Online, Gemini Enterprise
app also utilizes this zero-copy approach. This is a critical architectural distinction for
IT leaders concerned with compliance and storage costs:

Zero copying Identity first

Gemini Enterprise app retrieves information The platform supports Microsoft Entra ID 

directly from the source system at the 
 to manage user identities. This ensures that
moment of the query. It does not need to copy when a user searches for a contract in
your proprietary documents into a separate SharePoint, Gemini respects the exact user
Google search index to reason over them. permissions and security rules you have
already defined in Microsoft.

Our integration with Microsoft 365 showcases a The answers you receive are grounded in real-time
commitment to a secure and efficient architectural information, and your existing security rules are
approach. The zero-copy foundation of Gemini always respected. This not only enhances data
Enterprise app is a fundamental differentiator. sovereignty and reduces compliance risks but 

also prevents the creation of new silos.

13
 Contact sales

Example: 

Cross-platform workflow
Consider a Project Manager who needs to synthesize
data from Jira with a certain communication style found
in a Word document in Sharepoint for a status update.

The prompt

“Summarize the latest engineering risks from

the ‘Project X’ Jira board and draft an update
email to the steering committee based on the
Word doc titled ‘Communication Standards’
found in SharePoint.”

The execution

Gemini Enterprise app queries Jira (for the

risks) and SharePoint (for the style guide)
simultaneously. It then synthesizes this cross-
platform information into a cohesive draft.

The result

The user gets a ready-to-send email grounded

in real-time data from two different
ecosystems, achieved without logging into
multiple dashboards or manually copying and
pasting context.

14
 Contact sales

Security, data governance,

and policy controls
Security and governance
Security is a primary consideration for AI adoption. Our security promise is this:

Your data is yours Permissions are respected

Your prompts, inputs, and company data are When Gemini searches different systems, 

never used to train Google’s foundation models. it respects existing access controls. If a user
doesn’t have permission to see something,
Gemini won’t summarize it for them.

Gemini Enterprise app can ease some of the governance and

security challenges by acting as a single pane of glass for IT:

Centralized dashboard Model Armor Data sovereignty

IT admins can view, audit, This defense layer sits You can define where 

and manage all agents between the user and 
 your data is stored and
across the organization, the model. It proactively processed to meet regional
whether they were made 
 screens for malicious 
 regulatory requirements.
by Google, co-built with 
 inputs, like prompt injection
a partner, or created attacks, and filters outputs
individually by an employee. to prevent leaks of 

sensitive data.

15
 Contact sales

Governance at scale
As it is built on the Gemini Enterprise Agent Platform, and to scale AI securely,
Gemini Enterprise app offers enterprise-grade governance through 5 key areas:

Observability Agent Registry Tool Registry

Detailed logs, traces, and A single pane of glass for Centralized management 

metrics provide full admins to discover, manage, of tools and Model Context
transparency into agent and audit every agent Protocol (MCP) servers
invocations and data handling. operating across the available to agents.
organization.

Agent Identity Agent Gateway

Comprehensive visibility into agent roles, A central point of enforcement for policies
identities, and specific permissions they hold. governing agent-to-agent and agent-to-tool
interactions.

16
 Contact sales

Example 

use-cases
How does the Gemini Enterprise app
translate into day-to-day help for
specific departments? Let’s look at
how it can tackle friction points in
different scenarios.

17
 Contact sales

Enterprise search

Scenario The pain

An Account Executive is meeting with 
 Teams can spend hours looking for client
a high-stakes client and doesn’t have info across CRMs, internal folders, and
much time to prepare. public databases. This makes meeting
preparation labor-intensive and risks using
outdated information.

The solution Example prompt

Using Gemini Enterprise app to search for The Account Executive goes to Gemini
the necessary information gives you a Enterprise app and prompts: “My customer
unified view of your corporate data and is Cymbal Insurance and I’m meeting with
the web. By connecting to platforms like their Head of Procurement on Thursday.
Salesforce, SharePoint, and Google Drive, Find their latest earnings report and give
you get real-time, permissions-aware me a list of talking points.” Gemini
access to critical information faster and retrieves public financial filings and 

more reliably. cross-references internal CRM information
and service tickets to deliver a briefing
note instantly.

18
 Contact sales

Multimodal AI assistant

Scenario The pain

A Social Media Manager needs visuals 
 Teams have limited resources and often
for a product launch, but their design have to make more with less. That’s
teams are fully booked and they don’t 
 especially true for design and creative
have the budget for an external agency. needs, for which most people don’t have
dedicated budgets or resources.

The solution Example prompt

With Gemini Enterprise app’s multimodal Instead of waiting for a designer to
AI assistant, anyone can generate images, become free, or budget to be approved
videos, and reports from simple text for an agency, they prompt Gemini:
prompts. This empowers any employee 
 “Create a series of three images for a
to produce polished, brand-ready assets 
 social media campaign promoting our 

in seconds. new line of artisanal coffee beans.”

19
 Contact sales

The Deep Research agent

Scenario The pain

An executive at an automotive company 
 Strategic planning usually requires lots 

is evaluating a pivot in EV charging of “desk research”, like reading market
components. reports, analyzing competitor news, and
synthesizing trends. When done manually,
the process is slow, prone to bias, and can
keep you stuck in data-gathering mode.

The solution Example prompt

Deep Research slashes this time 
 The VP prompts Deep Research: “Analyze
from weeks to minutes. It doesn’t just the competitive landscape for EV charging
summarize search results; it explores in Northern Europe, focusing on recent
complex topics on your behalf, regulatory changes and competitor supply
synthesizing information across internal chain vulnerabilities.”
documents and external web sources into
comprehensive, cited reports.

20
 Contact sales
Contact sales

The Data Insights agent

Scenario The pain

A CFO notices an anomaly in Q3 global Complex spreadsheets are often where
travel expenses but can’t pinpoint the insights go to hide. Extracting a narrative
cause without a custom SQL query. from thousands of rows of raw data is
tedious, error-prone, and usually requires 

a data analyst who is already overbooked.

The solution Example prompt

The Data Insights agent allows employees The CFO uploads a raw expense ledger
to “talk” to their data. It can ingest raw files and asks the Data Insights agent to 

(CSVs, Sheets, Excel) and answer natural review it, uncover the top 5 expense
language questions. categories for Q3 compared to Q2, and
create a visualization of the analysis.

21
 Contact
Contactsales
sales

Agent Designer

Scenario The pain

A team of marketers are creating dozens Teams have repetitive and repeatable
of social media posts and emails for 
 processes that don’t justify working with 

a fast-moving campaign, but the tone 
 a developer to automate, but still eat up
varies significantly between writers. hours of time.

The solution Example prompt

With Agent Designer, a marketing manager The team lead uses Agent Designer to
can build a “campaign helper” simply by create a “brand voice reviewer.” While
typing instructions. You define the agent’s creating the agent, the lead uses the “brand
role (“You are a helpful editor”), give it guidelines” document to help set context.
access to specific files (“Use the brand They then prompt: “You are a strict editor.
guidelines PDF”), and share it with your Review all the text in the ‘social and email
team. No code required. copy’ document for our bold and
empathetic tone. Rewrite any passive voice
or jargon to match the guidelines.”

22
 Contact sales

Upskilling your
workforce with AI
Upskilling and training people in AI across an organization can be hard. To
succeed, organizations must consider their culture along with the technical
skills people need to thrive with AI.

This begins with the “human-in-the-loop” mindset. Employees are 


”editors-in-chief” who retain full accountability for their work. This way, AI
serves as a powerful engine to accelerate human creativity and judgment.

Traditional training programs might not meet this moment. A more effective
strategy involves micro-learning. Short and targeted educational modules
allow users to learn specific skills for immediate application.

Here are three pillars that can support a thriving culture of AI skills:

Central guidance Internal champions Measurement

An AI Council defines 
 Early adopters act as peer Leaders must track adoption
the governance and best coaches to guide 
 and sentiment, and devise
practices for the organization. their teams and share different learning strategies
success stories. for different levels of
proficiency and usage.

23
 Contact sales

Future-proof your AI
strategy with partners and
open industry standards
Deep partner ecosystem
No organization works with a single vendor, so Gemini Enterprise app is fundamentally
built on an open ecosystem philosophy, with a network of over 100,000 partners 

to ensure interoperability and prevent vendor lock-in. The objective is to let AI work
wherever organizational data is.

Open protocols
We're heading toward a future where different AI agents can work together to get
things done across different organizations. For this to happen, these tools need a safe
way to handle business and payments.

Google is pioneering the Agent-to-Agent (A2A) protocol and has co-developed the
Agent Payments Protocol (AP2) with financial leaders like American Express and
Mastercard. Adopting Gemini Enterprise app means building on a platform designed
for this future of open, interoperable commerce.

The Agent2Agent protocol

The Agent-to-Agent protocol (A2A) ensures that an agent built in Gemini can communicate
and collaborate with an agent built on another stack. This matters because you might build a
“customer service agent” in Gemini that needs to work with another agent running in, say,
Salesforce. The A2A protocol allows these agents to handshake, share context, and execute
the workflow securely. For an agent to truly take valuable action, like booking travel,
procuring software licenses, or restocking inventory, it needs some spending authority.

24
 Contact sales

The Agent Payments Protocol

The Agent Payments Protocol (AP2) is an open-source standard developed by Google in
collaboration with financial leaders, including American Express, Mastercard, PayPal, and
Adyen. AP2 introduces verifiable credentials called “mandates” that act as secure digital
contracts between you, the agent, and the merchant.

Intent mandates Cart mandates Payment mandates

This allows you to give an For supervised tasks, the A separate credential 

agent a specific mission agent builds a “shopping allows payment networks 

with strict guardrails. cart” and presents it to you. to identify that an AI Agent
When you approve, a “cart initiated the transaction,
mandate” is signed, locking enabling specialized fraud
in the exact items, prices, detection and risk
and other details so that management protocols.
what you approve is exactly
what gets purchased.

By adopting a platform built on open standards like A2A and AP2, you ensure 

your agents can work within the digital economy, not just in one vendor’s ecosystem.

25
 Contact sales

AI that delivers at all

levels, by Google
Google offers a complete architecture spanning five layers, from the chips and
infrastructure up to the models and agents. Google Cloud is the only provider that
researches, engineers, and deploys all five of those layers in-house. We co-design the
components to work as one system, and that deep integration provides an advantage. 

It allows you to deliver differentiated AI experiences with velocity, precision, and cost-
efficiency at scale, all on a foundation of uncompromising security.

At the base of it all is our AI Hypercomputer. Our supercomputing infrastructure gives

you access to the latest TPUs, CPUs, and NVIDIA GPUs, integrated with open software
and orchestration tools. All of this is designed to simplify AI infrastructure management,
push the limits of performance, and ultimately deliver more intelligence per dollar.

Gemini Enterprise
Agent Platform
and BigQuery
Gemini Enterprise app can easily connect
with BigQuery and BigLake, to better and
more accurately activate data. And you can
build custom agents using the Agent
Development Kit (ADK) in Gemini Enterprise
Agent Platform that can be governed in
Gemini Enterprise app. The ADK provides
enterprise blueprints to safely create
custom agents in Gemini Enterprise Agent
Platform using the models of your choice 

for your industry and business.

26
 Contact sales

Gemini Enterprise app and

Google Workspace

Gemini Enterprise app is built to connect to the apps your employees already use, including

Microsoft 365, Salesforce, and more. However, organizations can unlock transformative AI

value for individuals, teams, and the organization at large by connecting Gemini Enterprise

app with Google Workspace (Gmail, Drive, Docs, Sheets, Slides, and more).

Gemini is integrated directly into the Google Workspace apps. You can speed up the creation

of documents and presentations with AI drafts that understand your tone and business

context. You can also get useful AI meeting notes that are actionable and insightful. What’s

more, your work continues to inform the assistance you get across both Workspace and

Gemini Enterprise app, becoming even more personalized and helpful over time.

This is possible due to contextual awareness. Because agents can reason and share memory,

you don't need to re-prompt every time as you switch from one tool to the next:

If you define a strategy in Notes taken by Gemini in Gemini will share them with

Docs, an agent in Gmail will Google Meet will be turned the people in the meeting

know the context when you into nuanced suggestions in a Doc.

ask it to draft an update to and action items.

the team.

Presentations can be created with relevant Gemini in Drive can tell you anything you

sources from Drive. These presentations 
 need to know about your files, and it can also

can then be turned into AI-powered videos create audio overviews and summaries of

with Vids. your documents.

27
 Contact sales

03
Getting started

🍌 Create image

281
 Contact sales

Partner with Google

With Google Cloud, you get the advantage of a full AI stack. And to give you
flexibility, we meet you where you are. We know many organizations live in a
multi-vendor, multi-cloud world. Our platform is open by design, giving you the
freedom to innovate with our tools without being trapped by them.

Think of it as a “yes, and” strategy: Yes, our stack is engineered to work

perfectly together. AND, each layer is designed to plug into what you already
have. So whether you choose to go all-in on Google or adopt a multi-cloud
approach, you keep the freedom to adapt your strategy in the future without
the penalty of lock-in.

Finally, innovation cannot come at the cost of stability. This is why we provide a
foundation built on the same zero-trust security, sovereignty, and scale
reliability that keeps Google Search and YouTube online for billions of users.

Don't let your data stay trapped in silos and your people waste time and energy.
It’s time to stop stitching disconnected tools together and adopt a unified
system of intelligence.

Contact our Sales team

today to explore
solutions, set up a trial,
or learn more about the
Gemini Enterprise app.

29