Agentic Infrastructure for the Autonomous Enterprise on Google Cloud

Overview

This course provides a technical guide to enable Solution Architects to shift from building isolated chatbots to deploying persistent, Gemini Enterprise-enabled AI workers on Google Cloud. Participants will master agentic memory design, API-driven tool orchestration, and infrastructure governance using the Google Cloud Agent Platform—including the Vertex AI Reasoning Engine for persistent state management and Agent Extensions for departmental integration. Learners will move beyond ’Instructional Hope’ to technical enforcement, building the ’Paved Road’ required to orchestrate multi-agent fl eets and secure non-human identities.

Prerequisites

We recommend that attendees of this course have:

• Completion of Organising for AI Success
• Foundational knowledge of Google Cloud (VPC, IAM, Cloud Run)
• Helpful to be familiar with Python basics, REST API structures, and RAG concept

Target Audience

This course is designed for:

• Enterprise Architects who need to build a scalable, secure backbone for persistent, Gemini Enterprise-enabled AI workers
• Systems Integrators tasked with connecting AI to legacy ERP/CRM data via secure, cross-departmental tool orchestration using the Google Cloud Agent Platform
• IT Directors who need to move from ’Project Thinking’ to ’Platform Thinking’ to manage a global fleet of autonomous agents while maintaining governance and scalability

Objectives

By the end of this course, learners will be able to:

• Diagnose agency gaps by identifying conversational habits and mapping structural frictions
• Architect persistent, context-aware reasoning loops using the Build Tier
• Enforce Zero Trust patterns and runtime protections via the Govern Tier
• Mitigate reasoning drift and measure ROI using the Optimise & Scale Tiers
• Deploy persistent, Gemini Enterprise-enabled AI workers on Google Cloud
• Design agentic memory and tool orchestration
• Govern infrastructure using the Google Cloud Agent Platform

Outline

The Shift to Agency

• The pillars and governance
• The architectural friction forces
• The autonomous maturity scale

Building the ’Paved Road’

• Reference stack and tool archetypes
• The memory decision guide
• Multi-agent orchestration patterns
• The paved road lifecycle

The Autonomous Perimeter

• Threat modelling for agentic systems
• Identity hierarchy and credentials
• Defending the boundary: Model Armor
• Responsible AI and human-in-the-loop

Sustaining Autonomy

• Infrastructure ROI
• The GenAIOps lifecycle
• The innovation harvest

Summary and Quiz

• Review of core concepts

Exams and Assessments

There is no certification or exam associated with this course. The course includes scenario-based multiple choice questions to evaluate understanding of core concepts.

Hands-On Learning

Learners will engage in use cases, technical demos, practitioner artefacts, and facilitated discussions to apply concepts in practical scenarios. Activities include procurement reconciliation, model armour defence, global compliance scaling, forensic trace analysis, model armour configuration, structured audits, and addressing architectural friction and risk appetite.