Learn to deploy, automate, monitor, and debug machine learning and generative AI in production on Azure. 4-day official course covering MLOps and GenAIOps, aligned to Exam AI-300.

👥 1950 Enrolled ⏱️ 4 Days 💼 Intermediate Level ⭐ 5 ( 195 ) Reviews

500+Courses

10+ YrsExperience

95K+Global Learners

Live Online (Public Batch)

Instructor-led group class, fixed schedule

USD 2295

Private Group

Train your whole team, on your dates

Custom Quote

For Teams

Proceed to Buy Request Info

📄 Download Course PDF

100% Satisfaction Guarantee · Guaranteed to Run · Secure Payment

Course Overview

Training a model is a project. Keeping models and AI applications running reliably in production is a job, and it is a different one. Microsoft built AI-300T00 for the people doing that job, and the certification it leads to replaces the retired Azure Data Scientist Associate path.

The course covers what Microsoft calls AI operations, or AIOps, which spans two disciplines. MLOps handles traditional machine learning models: experimentation, automated training, hyperparameter tuning, pipelines, and CI/CD. GenAIOps handles generative AI applications and agents: prompt versioning, structured evaluation, monitoring, and tracing. Over four days you work through both.

The first half runs on Azure Machine Learning. You will preprocess data and configure featurization, run automated ML experiments, evaluate and compare models, configure MLflow for tracking, and assess models with the Responsible AI dashboard. From there you tune hyperparameters with sweep jobs, build components and pipelines for scalable workflows, then wire the whole thing into GitHub Actions for model training, feature-based development, environment management, and deployment.

The second half moves to Microsoft Foundry and generative AI. You plan a GenAIOps solution and select the right model, apply version control to prompts and agents through GitHub, design structured evaluation experiments with consistent scoring rubrics, and automate evaluations with batch Python and GitHub Actions. It closes on the two things that decide whether a generative AI application can be supported in production: monitoring, covering which metrics matter and how to interpret them, and tracing, covering what to trace, how to debug complex workflows, and how to make decisions from trace data.

Mode of Training

🏫 Classroom 💻 Live Online 🧪 Blended 👨‍👩‍👧‍👦 Private Group

Upcoming Schedules

Start Date	Time	Duration	Mode	Price
Sep 22, 2026	9:00 am - 5:00 pm	4 Days	online	$2295	GTR
Oct 06, 2026	9:00 am - 5:00 pm	4 Days	online	$2295	GTR
Oct 20, 2026	9:00 am - 5:00 pm	4 Days	online	$2295	GTR
Nov 03, 2026	9:00 am - 5:00 pm	4 Days	online	$2295	GTR

+ View more schedules

What you will learn

Implement end-to-end MLOps workflows using Azure Machine Learning
Automate model training, experimentation, and hyperparameter tuning
Build and manage machine learning pipelines for scalable workflows
Integrate GitHub Actions to enable CI/CD automation for ML models
Deploy, monitor, and manage machine learning and generative AI applications in production
Evaluate, optimize, and debug AI systems using GenAIOps practices, monitoring, and tracing

Who Should Attend This Course?

AI engineers and ML engineers responsible for models in production
Data scientists who need to operationalize their own work rather than hand it off
DevOps and platform engineers asked to own the AI pipeline
Cloud engineers building toward generative AI operations roles
Anyone whose DP-100 certification has retired and who needs the current equivalent

Prerequisites

Required:

Programming experience with Python
Experience developing and training machine learning models
A basic understanding of machine learning concepts

Recommended:

Familiarity with Microsoft Azure fundamentals
Basic knowledge of Azure Machine Learning
An understanding of DevOps practices including Git, source control, and CI/CD
Experience with command-line tools such as the Azure CLI
A basic understanding of generative AI concepts

The training requirement is real. This course teaches you to operationalize models, not to build them, so it assumes you have already trained and evaluated a model yourself.

📞 Talk to a Learning Advisor

Exam and Certification

This course prepares you for Exam AI-300: Operationalizing Machine Learning and Generative AI Solutions.

The exam validates the ability to set up infrastructure for machine learning operations and generative AI operations on Azure, covering secure and scalable MLOps infrastructure, automated provisioning and deployment with GitHub Actions, Bicep, and the Azure CLI, training orchestration, model registration and versioning, production monitoring, deploying generative AI solutions with Microsoft Foundry, quality assurance and safety evaluations, and optimizing RAG pipelines and fine-tuned models for performance, accuracy, and cost.

A note on the credential name: Microsoft's announcement introduced this as the Machine Learning Operations (MLOps) Engineer Associate certification, while the official study guide titles it Operationalizing Machine Learning and Generative AI Solutions. Both names are in circulation. The exam code AI-300 is the reliable identifier when booking.

The exam is booked and paid for separately from the course.

Important: the DP-100 transition

AI-300 replaces DP-100. Exam DP-100 and the Microsoft Certified: Azure Data Scientist Associate certification retired on 1 June 2026, along with the renewal assessment.

If you hold Azure Data Scientist Associate, it remains valid until its expiration date but cannot be renewed. AI-300 is how you hold a current credential in this area afterwards.

The scope has widened considerably. DP-100 validated designing and implementing data science solutions: data exploration, model training, evaluation, and deployment. AI-300 keeps training and evaluation but places far more weight on automation, infrastructure as code, CI/CD, lifecycle governance, observability, drift detection, cost control, and operationalizing generative AI systems.

That shift is the point. The industry moved from "can you build a model" to "can you run one."

📘 AI-300T00: Operationalize Machine Learning and Generative AI Solutions Outline

Experiment with Azure Machine Learning. Preprocess data and configure featurization, run an automated machine learning experiment, evaluate and compare models, configure MLflow for model tracking in notebooks, train and track models, and evaluate with the Responsible AI dashboard. Exercise: find the best classification model with Azure Machine Learning.

Hyperparameter tuning. Define a search space, configure a sampling method, configure early termination, and use a sweep job. Exercise: run a sweep job.

Pipelines. Create components, create a pipeline, and run a pipeline job. Exercise: run a pipeline job.

Trigger jobs with GitHub Actions. Understand the business problem and solution architecture, then use GitHub Actions for model training. Exercise included.

Feature-based development. Trigger workflows from feature branches. Exercise included.

Environments in GitHub Actions. Set up and manage environments across the deployment path. Exercise included.

Deploy a model with GitHub Actions. Model deployment within a CI/CD workflow. Exercise included.

Plan and prepare a GenAIOps solution. Explore use cases, select the right generative AI model, understand the development lifecycle of a language model application, and review the available tools and frameworks. Exercise: compare language models from the model catalog.

Manage prompts for agents in Microsoft Foundry with GitHub. Apply version control to prompts, understand Foundry agents and prompt versioning, organize prompts in GitHub repositories, and develop safe prompt deployment workflows. Exercise: develop prompt and agent versions.

Evaluate and optimize agents through structured experiments. Design evaluation experiments, apply Git-based workflows to optimization, and use evaluation rubrics for consistent scoring. Exercise: evaluate and compare AI agent versions.

Automate AI evaluations. Understand why automated evaluations matter, align evaluators with human criteria, create evaluation datasets, implement batch evaluations with Python, and integrate evaluations into GitHub Actions. Exercise: set up automated evaluations.

Monitor your generative AI application. Why monitoring matters, the key metrics to watch, monitoring with Azure, integrating monitoring into your app, and interpreting results. Exercise: enable monitoring for a generative AI application.

Analyze and debug with tracing. Why tracing matters, what to trace in generative AI applications, implementing tracing, debugging complex workflows with advanced patterns, and making decisions from trace data. Exercise: enable tracing for a generative AI application.

❓ Frequently Asked Questions

DP-100 and the Azure Data Scientist Associate certification retired on 1 June 2026, along with the renewal assessment. AI-300 is the successor. If you were studying for DP-100, registrations have closed. Your training and evaluation knowledge carries over, but you will need to add automation, CI/CD, infrastructure as code, monitoring, and generative AI operations.

MLOps applies to traditional machine learning models: training pipelines, model registration, versioning, drift detection. GenAIOps applies to generative AI applications and agents, where the artifacts are prompts, evaluations, and traces rather than trained weights. Microsoft groups both under AI operations, and this course teaches them as two halves of one job, which reflects how most teams now work.

Not necessarily, but you do need to have trained models before. The prerequisites ask for Python and experience developing and training machine learning models. DevOps engineers with that background often do very well here, since the second half of each section is CI/CD, environments, and automation.

Because that is where the operational work actually happens. The course uses GitHub Actions for triggering training jobs, feature-based development, environment management, model deployment, prompt versioning, and automated evaluations. If your organization uses Azure DevOps instead, the patterns transfer, but the labs are GitHub-based.

AI-103 is about building AI applications and agents in Microsoft Foundry. AI-300 is about operating them: versioning prompts, running structured evaluations, monitoring in production, and tracing failures. Teams often need both skill sets, and some people take both courses, but they are different roles.

The exam launched in beta in March 2026 and became generally available shortly after. Check current availability and any beta status on Microsoft Learn before booking, since new certifications sometimes have a gap before the official practice assessment is released.

Still have questions?

Reach out to our learning advisors for personalized guidance on choosing the right course, group training, or enterprise packages.

📞 Talk to an Advisor

What You Get with Microtek Learning

Instructor-Led Excellence

✓ Certified Instructor-led Training
✓ Top Industry Trainers
✓ Official Student Handbooks

Measurable Learning Outcomes

✓ Pre- & Post-Training Assessments
✓ Practice Tests
✓ Exam-Oriented Curriculum

Real-World Skill Building

✓ Hands-on Activities & Scenarios
✓ Interactive Online Courses
✓ Peer Collaboration (Not in self-paced)

Full Support & Perks

✓ Exam Scheduling Support ^*
✓ Learn & Earn Program ^*
✓ Support from Certified Experts
✓ Gov. & Private Pricing ^*

Our Clients

For over 10 years, Microtek Learning has helped organizations, leaders, students and professionals to reach their maximum potential. We have led the path by addressing their challenges and advancing their performances.