Google Cloud Professional Machine Learning Engineer Cert Prep
7h 9mAdvanced2025-03-13
Authors
Noah Gift
MLOps Expert | Solopreneur | Author | Adjunct Professor | CTO
Course details
Earning the Google Cloud Professional Machine Learning Engineer certification confirms that you’re able to build, evaluate, productionize, and optimize AI solutions by using Google Cloud capabilities and knowledge of conventional ML approaches. In this course, Noah Gift prepares you for the certification, starting with an overview of the exam—including the format of the exam, the time it should take, and how and where you can take the exam. Noah then dives into the six sections of the exam, covering what you need to know about: architecting low-code ML solutions; collaborating within and across teams to manage data and models; scaling prototypes into ML models; serving and scaling models; automating and orchestrating ML pipelines; and monitoring ML solutions.
Skills covered
Google Cloud PlatformMachine LearningSoftware Development ToolsGoogleCloud PlatformsCert PrepArtificial Intelligence (AI)Cloud ComputingSoftware Development
Concepts
0. Introduction
- 01 - Course and Google Professional Machine Learning Engineer exam overview
- 02 - Framing ML problems - Key terminology
1. Translating Business Challenges into ML Use Cases
- 03 - Building AI-enabled workflows
- 04 - Using AI tools to build AI tools
- 05 - Teaching MLOps at scale with GitHub
2. Defining ML Problems
- 06 - Simulations vs. experiment tracking
- 07 - When to use ML
- 08 - Supervised vs. unsupervised ML
- 09 - Optimization
- 10 - Clustering
3. Defining Business Success Criteria
- 11 - Defining business success criteria
4. Identifying Risks to Feasibility of ML Solutions
- 12 - MLOps hierarchy of needs
- 13 - Hidden costs of bespoke systems
- 14 - Data poisoning
5. Conclusion - Framing ML Problems
- 15 - Framing ML problems - Next steps
6. Introduction - Architecting a ML Solution
- 16 - Architecting a ML solution - Overview
- 17 - Architecting a ML solution - Key terminology
- 18 - Cloud developer workspace advantage
7. Designing a Reliable, Scalable, and Highly Available ML Solution
- 19 - What is continuous delivery
- 20 - Containerized ML microservices
- 21 - SRE mindset for MLOps
- 22 - Reproducible workflow
- 23 - Learn continuous integration
8. Choosing Appropriate Google Cloud Hardware Components
- 24 - Selecting heavy vs. light MLOps
- 25 - Key components of MLOps landscape
- 26 - Feature store vs. data warehouse
- 27 - Compute choice
9. Conclusion - Architecting a ML Solution
- 28 - Architecting a ML solution - Next steps
10. Introduction - Designing Data Preparation and Processing Systems
- 29 - Designing data preparation and processing systems - Overview
- 30 - Designing data preparation and processing systems - Key terminology
- 31 - Onboard to GCP
11. Exploring Data
- 32 - What is Google Colab
- 33 - Exploratory data analysis for life expectancy
- 34 - Data science setup with virtualenv and pip on Windows
- 35 - Graphing data for exploratory data analysis
12. Building Data Pipelines
- 36 - Labeling data
- 37 - Mechanical Turk labeling
- 38 - Cleaning up data
- 39 - Scaling data
- 40 - BigQuery data pipelines with Colab
13. Creating Input Features
- 41 - Feature engineering concepts
- 42 - Extracting features from public datasets
- 43 - Exploratory data analysis with Google BigQuery
14. Conclusion - Designing Data Preparation and Processing Systems
- 44 - Designing data preparation and processing systems - Next steps
15. Introduction - Developing ML Models
- 45 - Developing ML models - Overview
- 46 - Developing ML models - Key terminology
16. Building Models
- 47 - Using TensorFlow Playground
- 48 - Overfitting vs. underfitting
- 49 - Selecting the right metrics
17. Training Models
- 50 - Training models with TensorFlow and a GPU-enabled Docker
- 51 - Fine-tuning raw ingredients with Hugging Face
- 52 - Advantages of transfer learning
18. Scaling Model Training and Serving
- 53 - Operationalize microservices
- 54 - Monitoring and logging with Rust on Google App Engine
- 55 - Continuous integration using Rust with GitHub Actions
- 56 - Demo - Unit testing Rust
- 57 - Demo - GitHub copilot-enabled Rust
- 58 - Set up a GCP workstation with Python
- 59 - Demo - Google Cloud Shell
- 60 - Demo - Google Cloud Editor
- 61 - Demo - Google CLI SDK
- 62 - Demo - Google gcloud CLI
- 63 - Demo - Google App Engine Rust deployment
- 64 - Demo - Google App Engine Golang
19. Conclusion - Developing ML Models
- 65 - Developing ML models - Next steps
20. Introduction - Automating and Orchestrating ML Pipelines
- 66 - Automating and orchestrating ML pipelines - Overview
- 67 - Automating and orchestrating ML pipelines - Key terminology
21. Designing and Implementing Training Pipelines
- 68 - Prompt engineering for Google BigQuery with ChatGPT 4
- 69 - Getting started with Vertex AI
- 70 - Understanding TPUs
- 71 - TPUs as a technology transition
- 72 - Demo - TPU, PyTorch, and MNIST
22. Implementing Serving Pipelines
- 73 - TensorFlow serving with a GPU-enabled Docker
- 74 - A Rust and PyTorch microservice walkthrough
- 75 - Demo - Rust pretrained PyTorch microservice
23. Conclusion - Automating and Orchestrating ML Pipelines
- 76 - Automating and orchestrating ML pipelines - Next steps
24. Introduction - Monitoring, Optimizing, and Maintaining ML Solutions
- 77 - Monitoring, optimizing, and maintaining ML solutions - Overview
- 78 - Monitoring, optimizing, and maintaining ML solutions - Key terminology
25. ML Solutions
- 79 - Data drift explained by the naughty child problem
- 80 - Load testing with Locust
- 81 - Demo - Auditing via logs
- 82 - Demo - Logging dashboard
- 83 - Demo - Cloud web security scanner
- 84 - Demo - Querying logging output with BigQuery
- 85 - Demo - Load testing with Rust
- 86 - Five whys
- 87 - Using Google Courses
- 88 - Building a translator with Rust and Hugging Face
- 89 - Using PyTorch and Rust for stable diffusion
- 90 - Using Rust with PyTorch
- 91 - Building a CUDA GPU stress test
Conclusion
- 92 - Next steps