DATABRICKSDatabricks Certified Machine Learning Professional
The Databricks Certified Machine Learning Professional certification validates your ability to design, implement, and manage enterprise-scale machine learning solutions using advanced Databricks platform capabilities. It is for experienced ML engineers who build scalable pipelines, implement MLOps practices, and deploy production-ready models. Earning it demonstrates you can deliver comprehensive monitoring, testing, and deployment at scale.
230 practice questions · Updated 2026-07-30
MACHINE-LEARNING-PROFESSIONAL Curriculum
Every domain, objective, and concept the MACHINE-LEARNING-PROFESSIONAL exam measures.
- SparkML suitability assessment
- ML pipeline construction
- Estimator and transformer selection
- Model tuning with MLlib
- Model evaluation
- Batch and streaming scoring
- Inference type selection
- SparkML distributed training pipelines
- Pandas Function API for distributed training and inference
- Optuna distributed hyperparameter tuning
- Ray distributed hyperparameter tuning
- Vertical vs horizontal scaling trade-offs
- Model parallelism vs data parallelism
- Ray vs Spark for distributed ML training
- Nested Runs in MLflow
- Logging Custom Metrics, Parameters, and Artifacts
- Custom Model Objects with Real-Time Feature Engineering
- Point-in-Time Correctness
- Feature Lookup API for Training and Inference
- Automated Feature Pipelines with FeatureEngineering Client
- Online Tables Configuration with Databricks SDK
- Streaming Feature Ingestion
- On-Demand Features with Feature Serving
- Model lifecycle pipeline architecture
- Environment transition mechanisms
- Databricks features for lifecycle management
- Unit testing in Databricks notebooks
- Testing types across environment stages
- Integration testing for ML pipelines
- Organizing functions and unit tests
- Databricks Environment Design Principles
- Workspace and Compute Configuration
- Data Management for ML
- DABs Fundamentals
- Defining ML Assets with DABs
- Configuring Model Serving Endpoints
- Configuring MLflow Experiments
- Configuring Registered Models
- Deployment and Lifecycle Management
- Automated retraining triggers
- Data drift detection integration
- Performance degradation monitoring
- Retraining workflow orchestration
- Model selection strategy
- Model registry and versioning
- Champion-challenger evaluation
- Drift metrics table in Lakehouse Monitoring
- Lakehouse Monitoring table types and features
- Building monitors for different table types
- Components of monitoring pipelines
- Alerting mechanisms for drift thresholds
- Data drift detection methods
- Model performance trend evaluation
- Custom metrics in Lakehouse Monitoring
- Data granularity and feature slicing
- Endpoint health monitoring
- Blue-green deployment
- Canary deployment
- Comparing deployment strategies
- Databricks Model Serving rollout
- PyFunc model registration
- Custom model querying
- Custom model deployment
Percentages reflect share of the current practice bank, not official exam weightings — no structured per-skill weight is published for MACHINE-LEARNING-PROFESSIONAL, so none is invented.