| Topic | Details |
| Topic 1 | - Apply MLOps Practices: This domain targets the skills of Cloud Data Scientists and focuses on applying MLOps within the OCI ecosystem. It covers the architecture of OCI MLOps, managing custom jobs, leveraging autoscaling for deployed models, monitoring, logging, and automating ML workflows using pipelines to ensure scalable and production-ready deployments.
|
| Topic 2 | - OCI Data Science - Introduction & Configuration: This section of the exam measures the skills of Machine Learning Engineers and covers foundational concepts of Oracle Cloud Infrastructure (OCI) Data Science. It includes an overview of the platform, its architecture, and the capabilities offered by the Accelerated Data Science (ADS) SDK. It also addresses the initial configuration of tenancy and workspace setup to begin data science operations in OCI.
|
| Topic 3 | - Use Related OCI Services: This final section measures the competence of Machine Learning Engineers in utilizing OCI-integrated services to enhance data science capabilities. It includes creating Spark applications through OCI Data Flow, utilizing the OCI Open Data Service, and integrating other tools to optimize data handling and model execution workflows.
|
| Topic 4 | - Create and Manage Projects and Notebook Sessions: This part assesses the skills of Cloud Data Scientists and focuses on setting up and managing projects and notebook sessions within OCI Data Science. It also covers managing Conda environments, integrating OCI Vault for credentials, using Git-based repositories for source code control, and organizing your development environment to support streamlined collaboration and reproducibility.
|
| Topic 5 | - Implement End-to-End Machine Learning Lifecycle: This section evaluates the abilities of Machine Learning Engineers and includes an end-to-end walkthrough of the ML lifecycle within OCI. It involves data acquisition from various sources, data preparation, visualization, profiling, model building with open-source libraries, Oracle AutoML, model evaluation, interpretability with global and local explanations, and deployment using the model catalog.
|