What to study for each MLA-C01 domain
Feature Engineering and the Snowflake Feature Store
Covers designing entities and Feature Views, point-in-time correct retrieval, and serving features consistently for training and inference. It also addresses preventing leakage and training-serving skew.
Key skills
entities, Feature Views, Dynamic Tables, point-in-time joins, spine datasets, online and offline serving, leakage prevention
Model Development and the Model Registry
Covers developing models with Snowpark ML, versioning datasets, and logging models with metrics, signatures, and lineage. It also addresses evaluation, tuning, and promotion in the Model Registry.
Key skills
Snowpark ML, Model Registry, model versioning, signatures, Datasets, experiment tracking, hyperparameter tuning, evaluation metrics
Model Deployment and Serving
Covers deploying registered models for batch inference in SQL and real-time serving with Snowpark Container Services. It also addresses scaling, canary and shadow releases, and rollback.
Key skills
batch inference, model services, Snowpark Container Services, compute pools, canary deployment, shadow deployment, rollback
Model Monitoring and ML Observability
Covers monitoring production models for data drift, concept drift, and performance degradation in Snowflake ML. It also addresses baselines, alerting, segment analysis, and operational metrics.
Key skills
model monitors, data drift, concept drift, prediction drift, baselines, alert thresholds, segment analysis, endpoint observability
ML Governance, Security and Operations
Covers governing ML assets with RBAC, tags, masking, lineage, and audit in Snowflake Horizon. It also addresses pipeline automation, retraining operations, environments, and cost management.
Key skills
RBAC, least privilege, masking policies, tagging, Horizon lineage, Tasks, retraining pipelines, CI/CD, cost management