Details

---

Automating Your Business Processes With Google Antigravity

Duration: 2 Days

Introduction

In today's fast-moving corporate environment, professionals across every function continue to lose hours each week to repetitive manual work, consolidating spreadsheets, drafting routine reports, chasing approvals, and moving information between disconnected tools. The rise of agentic AI is changing this reality, shifting automation from a technical specialisation into a capability any business professional can direct, without writing a single line of code.

This 2-day hands-on course introduces participants to Google Antigravity, Google's agent-first automation platform built on Gemini 3, and shows how to direct autonomous agents to plan, execute, and verify real business workflows across the editor, browser, and Google Workspace. Participants will learn to translate everyday processes into clear agent instructions, review agent-generated Artifacts such as task lists and implementation plans, and apply Google Apps Script, generated and refined by the agent itself, to automate recurring tasks in Sheets, Docs, and Gmail. The course then extends this capability beyond the Google ecosystem, covering multiagent orchestration, multimodal automation of image, audio, and video content using tools such as Google AI Studio, and cross-application workflows spanning CRM and messaging platforms.

Throughout the programme, participants will be guided on the responsible use of agentic AI in the workplace, verifying agent outputs before acting on them, applying human-in-the-loop checkpoints for business-critical decisions, mitigating hallucination through proper grounding, and safeguarding data privacy when connecting Google Workspace accounts and third-party tools to automated workflows. Through structured, hands-on modules, participants will leave with the confidence to identify automation opportunities in their own roles, direct AI agents to build working solutions, and deploy sustainable, productionready automations that reduce manual workload across their teams.

Learning Outcomes

After attending this program, learners will be able to do the following:

• Explain how agentic AI differs from conventional generative AI tools and where Google Antigravity fits within the broader automation landscape
• Apply API, MCP, and CLI connection methods to link Antigravity agents with Google Workspace and other business data sources
• Utilise context, loop, and harness engineering principles to direct agents through multi step business tasks
• Develop a working Google Workspace automation that solves a genuine problem from the participant's own role
• Analyse multi agent orchestration patterns to determine the right structure for a given business process
• Construct automated workflows capable of processing image, audio, and video content alongside text
• Evaluate the accuracy, reliability, and governance compliance of agent generated outputs before deployment
• Design a scalable deployment plan that embeds version control and change management into everyday automation practice

Pre-Requisites

Participants must have an active Google AI Pro or Ultra subscription in place before the course begins, as the free tier does not provide sufficient token allocation to complete a full day of hands-on building in Google Antigravity. Each participant must come prepared with the following:

• Google AI Pro or Ultra subscription active before the course start date
• Google account with access to Gmail, Sheets, Drive, and Docs
• Google Antigravity desktop application installed and signed in

Key Content

Module 1: Understanding Agentic AI and the Google Antigravity Environment

• Explains what agentic AI is and how it differs from prompt based generative AI tools participants may already use
• Introduces how Google Antigravity applies the agentic AI paradigm across its Editor, Terminal, and Browser surfaces
• Guides participants through setting up instructions, memory, and configuration files that shape agent behaviour
• Covers the supported models available within Antigravity, including Gemini 3, Claude, and GPT OSS
• Explains how token usage and cost accumulate during agent runs and how to manage them responsibly
• Provides a first look at the Agent Manager and how Artifacts document what the agent has done

Module 2: Connecting Antigravity to Your Tools and Data

• Introduces API, MCP, and CLI as the three primary methods for connecting agents to external tools
• Walks participants through setting up Model Context Protocol servers for Google Workspace access
• Covers best practices in credentials management, including scoping access and storing API keys securely
• Explains how to identify which business tools already support these connection methods
• Provides hands on practice connecting Antigravity to Google Sheets, Drive, and Gmail
• Reinforces the habit of testing a connection with a small task before building a larger automation on top of it

Module 3: Prompting, Context, and Vibe Coding with Apps Script

• Builds a foundation in task level prompting so participants can direct agents clearly and consistently
• Introduces context engineering as a way of feeding agents the right background information for a task
• Explains loop engineering and how agents refine their own output through repeated review cycles
• Covers harness engineering as a way of setting boundaries and checks around what an agent is allowed to do
• Demonstrates vibe coding techniques for generating and refining Google Apps Script through natural language
• Gives participants hands on practice building and testing an Apps Script automation for Sheets, Docs, or Gmail

Module 4: Capstone Building Your Own Google Workspace Automation

• Guides participants through scoping a real, recurring task from their own role that is worth automating
• Encourages participants to apply the instructions, connections, and prompting skills covered across Modules 1 to 3
• Provides dedicated build time for participants to develop their own end to end Google Workspace automation
• Introduces a simple checklist for testing an automation before considering it complete
• Gives participants the opportunity to present their automation to the group

Module 5: Multi Agent Orchestration and Connecting External Systems

• Introduces the principles of multi agent systems and when a task calls for more than one agent
• Compares sequential, parallel, and approval gated workflows and the situations each is best suited for
• Explains how to apply human in the loop checkpoints for approval, review, and exception handling
• Extends the API, MCP, and CLI connection methods from Day 1 to non Google platforms
• Covers what to look for when assessing whether a new business tool can be safely connected to an agent
• Provides hands on practice connecting Antigravity to an external tool such as a CRM, Slack, or a REST API

Module 6: Automating Multimodal Workflows for Image, Audio, and Video

• Introduces multimodal AI capabilities and how agents process image, audio, and video inputs alongside text
• Explores Google AI Studio as an environment for testing multimodal prompts before building them into an automation
• Explains how to structure agent instructions so file handling stays reliable across different media types
• Covers file formats, storage considerations, and batching techniques that help avoid rate limits and processing errors
• Highlights common failure points when automations mix text and non text content in the same workflow
• Provides hands on practice building an automation that processes and routes a non text file, such as transcribing a recording or tagging product images

Module 7: Verifying, Evaluating, and Governing Automated Workflows

• Introduces methods for verifying agent outputs through Artifacts, screenshots, and browser recordings
• Covers how to define evaluation metrics such as accuracy, consistency, and correct fallback handling
• Walks participants through building a small test dataset to check an automation before relying on it
• Explains how hallucination risk can be reduced by grounding agents in verified sources and documents
• Covers data privacy considerations and the categories of information agents should never be given access to
• Introduces production habits such as documentation, error handling, retry logic, and modular workflow design

Module 8: Capstone Deploying and Scaling Automation Across the Organisation

• Covers how to schedule and deploy automations so they run reliably without constant supervision
• Introduces version control and change management practices suited to business critical workflows
• Provides dedicated build time for participants to develop an end to end automation spanning external systems
• Encourages participants to incorporate a multimodal component into their capstone build
• Gives participants the opportunity to present their completed automation in a live capstone session
• Closes with a discussion on building a roadmap for scaling agentic automation adoption across teams

Target Audience

• Operations and business process leads seeking to automate manual workflows.
• Executives and team leads exploring agentic AI for departmental efficiency
• Administrative, HR, and finance professionals managing repetitive Google Workspace tasks
• Cross-functional teams seeking to reduce manual data entry and reporting

Methodology

• PowerPoint Presentation
• Hands-on Demonstration
• Group Discussions
• Case Simulations
• Live Capstone Presentations