🔒 Guided

Pre-launch preview. Authorised access only.

Incorrect code

Guided by A Guide to Cloud
Explore AB-900 AI-901
AI-300

MLOps Engineer
Associate (Beta)

The new MLOps & GenAIOps certification replacing DP-100. Azure Machine Learning, MLflow, Microsoft Foundry, training pipelines, model deployment, evaluation, and RAG optimization — with Python code examples and real-world scenarios.

📖

Study Guide

25 interactive modules

5 free modules across all domains (1 per domain) + 20 practice questions. No account needed.

Start Free →
✍️

Practice Exam

200 exam-style questions

Study mode with explanations + timed exam simulation. 20 free questions.

Try Free Questions →

25

Modules

200

Questions

5

Domains

Free

Domain 0 included

📝 Sample Question

A data scientist runs 200 hyperparameter trials. Each trial takes 45 minutes on an A100 GPU with a limited compute budget. Which sampling algorithm converges fastest?

A. Grid search
B. Random sampling
C. Bayesian sampling ✓

📊 What you'll cover

Design and Implement an MLOps Infrastructure 5 modules
Implement ML Model Lifecycle and Operations 8 modules
Design and Implement a GenAIOps Infrastructure 5 modules
Implement GenAI Quality Assurance and Observability 4 modules
Optimize Generative AI Systems and Model Performance 3 modules

What you'll learn

🔧

Azure ML & MLflow

Workspaces, compute, experiment tracking, training pipelines — the full MLOps toolkit.

🤖

Microsoft Foundry

Deploy foundation models, manage prompts, configure GenAIOps infrastructure at scale.

📊

Evaluation & Monitoring

Groundedness, relevance, safety metrics. Monitor latency, cost, and quality in production.

🎯

RAG & Fine-Tuning

Chunking strategies, hybrid search, embedding models. The critical fine-tune vs RAG decision.

Full Curriculum

25 interactive modules across 5 exam domains

1

Design and Implement an MLOps Infrastructure

5 modules

Premium

Build and secure Azure ML workspaces. Manage data, environments, and compute. Deploy infrastructure with Bicep and GitHub Actions. Configure networking and CI/CD for ML projects.

  1. 1ML Workspace Setup
  2. 2Data, Environments & Components
  3. 3Compute Targets
  4. 4Infrastructure as Code
  5. 5Git & CI/CD for ML
Preview Domain
2

Implement ML Model Lifecycle and Operations

8 modules

Premium

Track experiments with MLflow. Automate training with AutoML, sweep jobs, and pipelines. Distribute training across GPUs. Register, version, deploy, and monitor models in production.

  1. 1MLflow Tracking
  2. 2AutoML & Hyperparameter Tuning
  3. 3Training Pipelines
  4. 4Distributed Training
  5. 5Model Registration & Versioning
  6. 6Responsible AI Gates
  7. 7Deploying to Endpoints
  8. 8Drift, Monitoring & Retraining
Preview Domain
3

Design and Implement a GenAIOps Infrastructure

5 modules

Premium

Set up Microsoft Foundry hubs and projects. Configure network security. Deploy and version foundation models. Implement PromptOps — design, compare, version, and ship prompts.

  1. 1Foundry: Hubs & Projects
  2. 2Network Security & IaC
  3. 3Deploying Foundation Models
  4. 4Model Versioning Strategies
  5. 5PromptOps: Design to Ship
Preview Domain
4

Implement GenAI Quality Assurance and Observability

4 modules

Premium

Evaluate GenAI with quality metrics — groundedness, relevance, coherence, fluency. Configure safety evaluations. Monitor latency, throughput, and costs in production.

  1. 1Evaluation Datasets & Metrics
  2. 2Safety & Custom Metrics
  3. 3Monitoring GenAI
  4. 4Cost, Logging & Debugging
Preview Domain
5

Optimize Generative AI Systems and Model Performance

3 modules

Premium

Optimise RAG with chunking and hybrid search. Select and tune embedding models. Master fine-tuning — methods, synthetic data, and the critical decision matrix.

  1. 1RAG Optimization
  2. 2Embeddings & Hybrid Search
  3. 3Fine-Tuning: Methods to Production
Preview Domain

Practice Exam Lab

200 original questions — two study modes

📖

Study Mode

Learn as you go

  • ✓ See explanation after each question
  • ✓ "Why wrong" for every option
  • ✓ Real-world context & exam tips
  • ✓ Microsoft Learn links
⏱️

Exam Mode

Simulate the real thing

  • ✓ 100 minutes timed session
  • ✓ Randomised question order
  • ✓ Score breakdown by domain
  • ✓ Pass/fail against 700 / 1000
1

MLOps Infrastructure

35 questions · Exam weight: 15-20%

4 Free

10

Easy

17

Medium

8

Hard

Every question includes a scenario, detailed explanation, and a link to Microsoft Learn.

Preview Questions
2

ML Model Lifecycle

60 questions · Exam weight: 25-30%

4 Free

15

Easy

30

Medium

15

Hard

Every question includes a scenario, detailed explanation, and a link to Microsoft Learn.

Preview Questions
3

GenAIOps Infrastructure

45 questions · Exam weight: 20-25%

4 Free

12

Easy

23

Medium

10

Hard

Every question includes a scenario, detailed explanation, and a link to Microsoft Learn.

Preview Questions
4

GenAI Quality & Observability

30 questions · Exam weight: 10-15%

4 Free

8

Easy

15

Medium

7

Hard

Every question includes a scenario, detailed explanation, and a link to Microsoft Learn.

Preview Questions
5

Optimize GenAI Performance

30 questions · Exam weight: 10-15%

4 Free

8

Easy

15

Medium

7

Hard

Every question includes a scenario, detailed explanation, and a link to Microsoft Learn.

Preview Questions

🎁 Free tier: 20 questions. No account needed.

Choose your path

Start free. Upgrade when you're ready.

Practice Exam

$14

one-time purchase

  • ✓ 200 exam-style questions
  • ✓ Study mode + Exam mode
  • ✓ Detailed explanations

Study Guide

$19

one-time purchase

  • ✓ 25 interactive modules
  • ✓ Flashcards & ELI5
  • ✓ Progress tracking
Recommended

Complete Bundle

$29

save $4

  • ✓ Everything in both
  • ★ Best value

🎁 Free tier: 5 free modules across all domains (1 per domain) + 20 practice questions. No account needed. No account needed.

About the AI-300 Exam

Exam code
AI-300
Replaces
DP-100 (Azure Data Scientist Associate)
Duration
100 minutes
Passing score
700 / 1000
Question types
MCQ, multi-select, case studies, labs
Cost
$165 USD

Exam Domains & Weights

D1: MLOps Infrastructure (15-20%)
D2: ML Model Lifecycle (25-30%)
D3: GenAIOps Infrastructure (20-25%)
D4: GenAI Quality & Observability (10-15%)
D5: Optimize GenAI Performance (10-15%)
View on Microsoft Learn →

Frequently Asked Questions

What is AI-300?
AI-300 is the MLOps Engineer Associate certification — an intermediate exam covering Azure Machine Learning, MLflow, Microsoft Foundry, and the full ML/GenAI operational lifecycle.
Does AI-300 replace DP-100?
Yes. DP-100 (Azure Data Scientist Associate) is being replaced by AI-300, which adds GenAIOps, Foundry, and prompt management to the MLOps curriculum.
Is AI-300 still in beta?
Yes — AI-300 launched as a beta exam in March 2026. Content may change before GA (expected June-July 2026). We update our guide as the exam evolves.
What's in the free tier?
5 free modules (one per domain) across all 5 exam areas, plus 20 practice questions. No account needed.
Do I need Python experience?
Yes — AI-300 is an intermediate exam that assumes Python proficiency, basic ML knowledge, and familiarity with DevOps practices like Git and CI/CD.
Is this a one-time purchase?
Yes. Pay once, access forever. No subscription or recurring fees.

Ready to start?

Try the free tier. Upgrade when you're ready to pass.

Start Free → Try Practice Quiz

Guided

I learn, I simplify, I share.

A Guide to Cloud YouTube Feedback

© 2026 Sutheesh. All rights reserved.

Guided is an independent study resource and is not affiliated with, endorsed by, or officially connected to Microsoft. Microsoft, Azure, and related trademarks are property of Microsoft Corporation. Always verify information against Microsoft Learn.