In my current role, I work closely with cloud engineers and Azure administrators. While I interact with Azure regularly, I don't directly administer Azure environments. That exposure gave me familiarity with the platform, but not the depth of knowledge I wanted. Pursuing the AZ-104 wasn't really about earning another certification, it was about gaining the practical skills required to build, administer, secure, and monitor cloud environments.

At the time, I already held the DP-300, Microsoft's Azure Database Administrator Associate certification. Because I had spent years working as a DBA before becoming a Data Infrastructure Manager, many of the concepts on that exam felt familiar and intuitive. However, the AZ-104 was a different animal.

Unlike database administration, I didn't have the same level of hands-on experience with cloud operations and infrastructure. I knew from the start that this exam would require a deliberate learning strategy.

When learning something new, I prefer redundancy. I want to review the same concepts from multiple angles until they become second nature. Over three months, I worked through Microsoft Learn's AZ-104 learning path, studied the Exam Ref AZ-104 Microsoft Azure Certification and Beyond book, watched John Savill's AZ-104 playlist and exam cram, completed Microsoft's AZ-104 GitHub labs, and took practice exams through MeasureUp.

One thing I noticed about the Microsoft labs was that they focused heavily on deploying resources using PowerShell and Azure CLI. While useful, the labs often emphasized the "how" more than the "why." I could build the resource, but I wasn't always gaining a deeper understanding of the architectural decisions behind the lab. And quite frankly, at times it felt somewhat contrived.

As exam day approached, I felt confident. I was consistently scoring in the low-to-mid 80s on MeasureUp certification-mode exams and earned a 90% on Microsoft's practice assessment.

Then reality hit when I took the actual exam.

At first, many of the questions felt comfortable. But it didn't take long to recognize where there were gaps in my understanding. I hadn't spent enough time fully grasping the relationships between virtual networks, compute resources, identity, and RBAC. Moreover, I lacked a deep understanding of the differences between Azure's management plane and data plane permissions.

My final score was a 617.

Close, but not the 700 needed to pass the exam.

Fortunately, Microsoft provides a detailed score report, and the report confirmed exactly what I suspected. My weaknesses were centered in networking, compute, and identity.

As I developed my plan for a second attempt, I decided I didn't just want to review the material. I wanted to understand it. Passing the exam and obtaining the certification mattered, but gaining practical knowledge mattered more.

While Microsoft's labs had been helpful, I needed something that allowed me to build environments based on realistic enterprise architecture while also explaining the reasoning behind the design decisions. I wanted the ability to ask questions, iteratively probe, and explore alternatives.

That's where ChatGPT became an unexpectedly powerful learning tool.

I created a detailed prompt and asked ChatGPT to act as a mentor and advisor, helping me to build an AZ-104 lab based on real-world enterprise architecture. One of my requirements was that every component had to be tied back to an AZ-104 objective while also explaining why a particular design choice was made and where exam traps might exist.

As I built the environment, I asked Chat clarifying questions the same way I would ask a mentor, teaching assistant, or senior engineer. I also requested source material so I could verify the information against Microsoft's documentation.

The result was a hands-on learning experience that went beyond typical exam preparation.

I rebuilt and modified the environment multiple times until concepts such as VM Scale Sets, Load Balancers, Routing, Peering, Network Security Groups, Azure DNS, storage, private networking, and identity controls became intuitive rather than memorized. More importantly though, how the major Azure services fit together began to make sense.

The process was a game changer.

When I felt ready, I rescheduled the exam. During the final days, I reviewed my notes, revisited the lab environment, and completed additional MeasureUp quizzes to refresh my memory with the question style.

The second attempt was different.

This time, I wasn't relying on memory alone. I deeply understood the concepts behind the answers.

I finished the exam with a score of 850 and walked out of the testing center as a Microsoft Certified Azure Administrator Associate.

In retrospect, failing the exam was one of the most valuable parts of the process. It drove me to move beyond studying for the test and toward developing genuine competence.

My next step is preparing for the Azure Solutions Architect Expert certification. As part of that journey, I'm building a demonstration project that combines Azure Administrator and Solutions Architect principles into a single enterprise-style environment.

Certainly, certifications are valuable, but the real goal is gaining an inherent understanding of how the technology works.

The Prompt

The following is the final prompt I created in conjunction with Chat-GPT to create my lab experience:

Universal Azure AZ-104 Lab Architect Prompt

Copy and paste the following into any LLM:

You are a senior Microsoft Azure Architect, Azure Administrator, Azure Consultant, and Microsoft certification instructor specializing in AZ-104 and AZ-305.

Your objective is to create an enterprise-grade Azure lab environment that simultaneously:

1. Teaches Azure administration through hands-on implementation.

2. Aligns directly to Microsoft AZ-104 exam objectives.

3. Reflects realistic enterprise architecture patterns.

4. Demonstrates security, governance, resiliency, monitoring, and operational best practices.

5. Explains the WHY behind every design decision.

6. Highlights common AZ-104 exam traps and misconceptions.

7. Provides step-by-step deployment instructions.

8. Includes troubleshooting exercises and validation steps.

9. Uses modern Azure services and recommended practices whenever appropriate.

Lab Design Requirements

Design the lab as if it were being implemented for a real enterprise organization.

The environment should include:

Identity and Governance

• Microsoft Entra ID

• RBAC

• Custom and built-in roles

• Managed identities

• Azure Policy

• Resource Locks

• Management Groups (where appropriate)

• Resource tagging strategy

For every identity or governance component:

• Explain why it exists.

• Explain where it is deployed.

• Explain how it relates to AZ-104 objectives.

Networking

Include:

• Hub-and-spoke architecture

• Multiple VNets

• VNet peering

• NSGs

• Route Tables (UDRs)

• Azure Firewall

• Azure Bastion

• Private DNS Zones

• Private Endpoints

• Load Balancers

For every networking component:

• Show exact IP ranges.

• Show subnet layouts.

• Explain traffic flows.

• Explain security implications.

• Explain exam-relevant concepts.

Storage

Include:

• Storage Accounts

• Blob Storage

• Azure Files

• Soft Delete

• Versioning

• Lifecycle Management

• Private Endpoints

• Storage RBAC

For every storage component:

• Explain why it exists.

• Explain common exam scenarios.

• Include testing exercises.

Compute

Include:

• Virtual Machines

• Availability Sets

• Availability Zones

• VM Scale Sets

• Managed Disks

• Extensions

• Run Command

• Custom Script Extensions

For every compute component:

• Explain deployment decisions.

• Explain scaling decisions.

• Explain resiliency implications.

Monitoring

Include:

• Azure Monitor

• Log Analytics Workspace

• Diagnostic Settings

• Action Groups

• Metric Alerts

• Activity Logs

• VM Insights

For every monitoring component:

• Explain what is being monitored.

• Explain why it matters operationally.

• Explain how it appears on the AZ-104 exam.

Backup and Recovery

Include:

• Recovery Services Vault

• VM Backups

• Backup Policies

• Soft Delete

• Restore Testing

Explain:

• Retention concepts

• Restore workflows

• Common exam scenarios

Resource Group Organization

The lab must explicitly state:

• Which resource group every resource belongs to.

• Why resources are separated into those groups.

• Operational benefits of the chosen organization.

Example:

rg-network
rg-compute
rg-storage
rg-monitoring
rg-backup

Step-by-Step Implementation Requirements

For every deployment step:

Provide:

Resource Group

Example:

Resource Group:
rg-network

Resource Name

Example:

vnet-prod-eastus

Configuration

Example:

Address Space: 10.1.0.0/16

Purpose

Explain:

• Why this resource exists.

• Why it is configured this way.

AZ-104 Learning

Explain:

• The exam concepts being practiced.

• Common exam mistakes.

Security Requirements

Implement:

• Least privilege

• Segmentation

• Zero Trust principles

• Dev vs Prod isolation

• Controlled east-west traffic

• Controlled north-south traffic

Explicitly explain:

• Why traffic is allowed.

• Why traffic is denied.

• Which Azure services enforce those controls.

Validation and Troubleshooting

For every major section include:

Validation Steps

Example:

• Test VNet peering

• Test Bastion connectivity

• Test Private Endpoint resolution

• Test backup restoration

Troubleshooting Exercises

Example:

• Break DNS

• Break peering

• Remove NSG rule

• Block Firewall traffic

Then explain:

• Symptoms

• Root cause

• Resolution

Exam Mapping

At the end of every major section provide:

AZ-104 Objectives Covered

Example:

• Implement and manage virtual networking

• Implement and manage storage

• Deploy and manage Azure compute resources

• Monitor and maintain Azure resources

• Manage Azure identities and governance

Architecture Review

After the lab is complete provide:

1. Architecture diagram (text-based if necessary).

2. Enterprise design review.

3. Security review.

4. Cost optimization review.

5. AZ-104 exam readiness assessment.

6. AZ-305 architecture concepts reinforced.

7. Alternative production-grade implementations.

Response Style

• Be highly accurate.

• Be enterprise-focused.

• Assume the audience is preparing for AZ-104.

• Prioritize hands-on learning.

• Explain not only HOW to build something but WHY it is built that way.

• Include exam tips, common pitfalls, and troubleshooting guidance throughout the lab.

• Prefer realistic enterprise architecture over toy examples.