Exam Overview

# Databricks MLflow Certification - Exam Topics

exam_topics = {
    "MLflow Tracking (~30%)": {
        "skills": [
            "Create and manage experiments",
            "Log parameters, metrics, and artifacts",
            "Use autologging for popular ML frameworks",
            "Search and compare runs using the API",
            "Understand the tracking server architecture",
            "Query runs with mlflow.search_runs()"
        ],
        "key_apis": [
            "mlflow.start_run()",
            "mlflow.log_param() / mlflow.log_params()",
            "mlflow.log_metric() / mlflow.log_metrics()",
            "mlflow.log_artifact() / mlflow.log_artifacts()",
            "mlflow.autolog()",
            "mlflow.search_runs()"
        ]
    },
    "MLflow Models & Registry (~30%)": {
        "skills": [
            "Save and load models with different flavors",
            "Understand model signatures and input examples",
            "Register models in the Model Registry",
            "Transition models between stages",
            "Add descriptions and tags to registered models",
            "Use models URI scheme (models:/name/stage)"
        ],
        "key_apis": [
            "mlflow.sklearn.log_model()",
            "mlflow.pyfunc.load_model()",
            "mlflow.register_model()",
            "MlflowClient().transition_model_version_stage()",
            "mlflow.models.infer_signature()"
        ]
    },
    "MLflow Projects & Recipes (~20%)": {
        "skills": [
            "Define MLproject files with entry points",
            "Specify conda and Docker environments",
            "Run projects from Git repos or local dirs",
            "Understand MLflow Recipes (formerly Pipelines)",
            "Configure recipe profiles and steps"
        ],
        "key_apis": [
            "mlflow.projects.run()",
            "MLproject YAML structure",
            "conda.yaml environment spec",
            "mlflow recipes inspect / run"
        ]
    },
    "Model Deployment (~20%)": {
        "skills": [
            "Serve models locally with mlflow models serve",
            "Score models in batch with mlflow.pyfunc",
            "Deploy models in Docker containers",
            "Understand REST API input/output formats",
            "Deploy to cloud platforms (Databricks, SageMaker)"
        ],
        "key_apis": [
            "mlflow models serve --model-uri",
            "mlflow models build-docker",
            "mlflow.deployments",
            "mlflow.sagemaker.deploy()"
        ]
    }
}

# Study Plan for Databricks MLflow Certification

study_plan = {
    "Week 1: Foundations & Tracking": {
        "focus": "MLflow Tracking (~30% of exam)",
        "tasks": [
            "Install MLflow and set up a local tracking server",
            "Create experiments and log 10+ runs with parameters/metrics",
            "Practice autologging with scikit-learn and TensorFlow",
            "Use mlflow.search_runs() with filter strings",
            "Complete Lesson 2 of this course (with practice questions)",
            "Read MLflow Tracking official documentation"
        ]
    },
    "Week 2: Models & Registry": {
        "focus": "MLflow Models & Registry (~30% of exam)",
        "tasks": [
            "Log models with sklearn, pyfunc, and tensorflow flavors",
            "Create model signatures and input examples",
            "Register 3+ models in the Model Registry",
            "Practice stage transitions: None -> Staging -> Production",
            "Complete Lesson 3 of this course (with practice questions)",
            "Read MLflow Models and Registry documentation"
        ]
    },
    "Week 3: Projects, Recipes & Deployment": {
        "focus": "Projects (~20%) + Deployment (~20%)",
        "tasks": [
            "Create an MLproject file with conda environment",
            "Run a project from a Git repo with mlflow.projects.run()",
            "Serve a model locally and test the REST API",
            "Build a Docker image with mlflow models build-docker",
            "Complete Lessons 4 and 5 of this course",
            "Read MLflow Projects and Deployment documentation"
        ]
    },
    "Week 4: Review & Practice": {
        "focus": "Full review + practice questions",
        "tasks": [
            "Complete Lesson 6 (Exam Tips & Practice questions)",
            "Take timed practice quiz (45 questions, 90 minutes)",
            "Review all Quick Reference tables in Lesson 6",
            "Re-do any practice questions you got wrong",
            "Review the FAQ section in Lesson 6",
            "Register for the exam when you score 80%+ on practice"
        ]
    }
}

# Prerequisites checklist
prerequisites = {
    "Python": [
        "Comfortable with Python 3.x syntax",
        "Experience with pip and virtual environments",
        "Basic understanding of REST APIs"
    ],
    "Machine Learning": [
        "Understand the ML lifecycle (train, evaluate, deploy)",
        "Familiar with scikit-learn or similar frameworks",
        "Know what hyperparameters, metrics, and artifacts are",
        "Understand model serialization (pickle, joblib, ONNX)"
    ],
    "MLflow Basics": [
        "Installed and used MLflow locally",
        "Logged at least a few runs with parameters and metrics",
        "Used the MLflow UI to compare experiments",
        "Familiar with mlflow CLI commands"
    ],
    "Optional but Helpful": [
        "Experience with Databricks workspace",
        "Docker basics (for deployment questions)",
        "Git basics (for MLflow Projects questions)"
    ]
}