AWS ML Engineer Associate (MLA-C01) Study Plan¶

10-Week Comprehensive Study Schedule¶

Week 1: Data Ingestion and Storage¶

Day 1-2: S3 Data Lake Fundamentals¶

Study S3 as the primary data lake for ML workloads
Learn data formats for ML: Parquet, CSV, JSON, RecordIO, TFRecord
Understand S3 performance optimization: prefixes, multipart upload
Study S3 lifecycle policies for training data management
Hands-on: Create S3 data lake with structured and unstructured data
Review Notes: notes/01-data-preparation.md

Day 3-4: Data Ingestion Services¶

Study Kinesis Data Streams and Firehose for streaming data
Learn AWS Database Migration Service for database sources
Understand data catalog and metadata management
Hands-on: Set up Kinesis pipeline to S3 for real-time data
Lab: Configure Glue Crawler to catalog S3 data

Day 5-6: AWS Glue ETL¶

Study Glue ETL job types: Spark, Python Shell, Ray
Learn Glue Data Catalog and crawlers
Understand Glue Studio for visual ETL authoring
Study Glue DataBrew for visual data preparation
Hands-on: Build Glue ETL job to transform raw data
Lab: Use Glue DataBrew for data profiling and cleaning

Day 7: Week 1 Review and Practice¶

Complete data ingestion practice questions
Build end-to-end data pipeline
Review any weak areas identified

Week 2: Data Preparation with SageMaker¶

Day 8-9: SageMaker Data Wrangler¶

Study Data Wrangler import sources and connectors
Learn data transformation steps and custom transforms
Understand data visualization and analysis features
Study export options: Pipeline, Processing, Feature Store
Hands-on: Create Data Wrangler flow with transformations
Lab: Export Data Wrangler flow to SageMaker Pipeline

Day 10-11: SageMaker Processing¶

Study Processing job configuration and containers
Learn built-in containers: scikit-learn, Spark, custom
Understand distributed processing with Spark on Processing
Hands-on: Run Processing job for feature engineering
Lab: Implement custom processing container

Day 12-13: Feature Engineering Techniques¶

Study encoding methods: one-hot, label, target encoding
Learn scaling and normalization techniques
Understand feature selection and dimensionality reduction
Study handling missing data and outliers
Hands-on: Implement feature engineering pipeline
Review Notes: notes/01-data-preparation.md

Day 14: Week 2 Review and Integration¶

Complete data preparation practice questions
Design end-to-end feature engineering workflow
Practice data quality assessment scenarios

Week 3: Feature Store and Data Quality¶

Day 15-16: SageMaker Feature Store¶

Study Feature Store concepts: feature groups, records, features
Learn online store (low-latency reads) vs offline store (batch)
Understand feature group creation and ingestion
Study time-travel queries for point-in-time correctness
Hands-on: Create feature groups and ingest data
Lab: Query features for training and inference

Day 17-18: Data Quality and Validation¶

Study SageMaker Data Quality monitoring
Learn data profiling and schema validation
Understand data distribution checks and statistics
Study SageMaker Clarify for data bias detection
Hands-on: Set up data quality baseline and monitoring
Lab: Run Clarify bias analysis on training data

Day 19-20: Data Pipeline Integration¶

Study end-to-end data pipeline patterns
Learn pipeline orchestration with SageMaker Pipelines
Understand data versioning and lineage tracking
Hands-on: Build automated data preparation pipeline
Review Notes: notes/01-data-preparation.md

Day 21: Week 3 Review and Practice¶

Complete Feature Store and data quality practice questions
Design complete data preparation architecture
Practice data pipeline troubleshooting

Week 4: Model Development - Training¶

Day 22-23: SageMaker Training Jobs¶

Study training job configuration: instance types, input/output
Learn distributed training: data parallelism, model parallelism
Understand managed Spot training for cost optimization
Study checkpointing for training resumption
Hands-on: Run training jobs with different configurations
Lab: Implement distributed training with data parallelism
Review Notes: notes/02-model-development.md

Day 24-25: Built-in Algorithms¶

Study XGBoost for tabular classification and regression
Learn Linear Learner for linear models
Understand BlazingText for text classification and Word2Vec
Study Image Classification and Object Detection
Study K-Means, PCA, and other unsupervised algorithms
Hands-on: Train models with multiple built-in algorithms
Lab: Compare algorithm performance on same dataset

Day 26-27: Custom Training¶

Study Bring Your Own Container (BYOC) requirements
Learn Script Mode with framework containers
Understand TensorFlow, PyTorch, and Hugging Face on SageMaker
Study SageMaker Training Compiler for optimization
Hands-on: Train custom model using Script Mode
Lab: Build and push custom training container

Day 28: Week 4 Review and Integration¶

Complete training practice questions
Compare different training approaches
Practice instance type selection for training

Week 5: Model Development - Tuning and Evaluation¶

Day 29-30: Hyperparameter Tuning¶

Study SageMaker Automatic Model Tuning (AMT)
Learn search strategies: Bayesian, random, grid, hyperband
Understand warm start for tuning job continuation
Study early stopping for poorly performing jobs
Hands-on: Run hyperparameter tuning jobs
Lab: Compare tuning strategies on same model

Day 31-32: Model Evaluation and Experiments¶

Study evaluation metrics: accuracy, precision, recall, F1, AUC-ROC
Learn regression metrics: RMSE, MAE, R-squared
Understand SageMaker Experiments for tracking runs
Study SageMaker Debugger for training insights
Hands-on: Track experiments and compare models
Lab: Use Debugger to analyze training issues

Day 33-34: Model Registry and Autopilot¶

Study SageMaker Model Registry for versioning
Learn model package groups and approval workflows
Understand SageMaker Autopilot for AutoML
Study SageMaker JumpStart for pretrained models
Hands-on: Register models and set up approval workflow
Lab: Run Autopilot experiment and analyze results
Review Notes: notes/02-model-development.md

Day 35: Week 5 Review and Practice¶

Complete model development practice questions
Design model training and evaluation workflow
Practice model selection scenarios

Week 6: Deployment - Endpoints and Inference¶

Day 36-37: Real-Time Inference¶

Study SageMaker endpoint creation and configuration
Learn multi-model endpoints for cost optimization
Understand multi-container endpoints for pipeline inference
Study auto-scaling configuration for endpoints
Hands-on: Deploy model to real-time endpoint with auto-scaling
Lab: Set up multi-model endpoint
Review Notes: notes/03-deployment-orchestration.md

Day 38-39: Batch and Serverless Inference¶

Study Batch Transform for large-scale predictions
Learn Serverless Inference for intermittent traffic
Understand Asynchronous Inference for large payloads
Study inference optimization with SageMaker Neo
Hands-on: Run batch transform job on large dataset
Lab: Deploy serverless inference endpoint

Day 40-41: Deployment Strategies¶

Study blue/green deployments for endpoints
Learn canary deployments with traffic shifting
Understand shadow deployments for testing
Study deployment guardrails and rollback
Hands-on: Implement blue/green endpoint update
Lab: Configure canary deployment with auto-rollback

Day 42: Week 6 Review and Integration¶

Complete deployment practice questions
Compare inference options for different scenarios
Practice endpoint troubleshooting

Week 7: ML Pipelines and MLOps¶

Day 43-44: SageMaker Pipelines¶

Study Pipeline steps: Processing, Training, Tuning, Transform
Learn conditional steps and parallel execution
Understand pipeline parameters and caching
Study pipeline execution and monitoring
Hands-on: Build end-to-end ML pipeline
Lab: Add conditional logic and approval steps

Day 45-46: Step Functions and CI/CD¶

Study Step Functions for ML workflow orchestration
Learn SageMaker integration with Step Functions
Understand CodePipeline for ML CI/CD
Study SageMaker Projects and MLOps templates
Hands-on: Create Step Functions ML workflow
Lab: Set up CI/CD pipeline for model deployment

Day 47-48: MLOps Best Practices¶

Study infrastructure management for ML
Learn container management with ECR
Understand model lineage tracking
Study SageMaker Model Cards for documentation
Hands-on: Implement complete MLOps workflow
Review Notes: notes/03-deployment-orchestration.md

Day 49: Week 7 Review¶

Complete MLOps practice questions
Design automated ML pipeline architecture
Practice pipeline troubleshooting scenarios

Week 8: Monitoring and Maintenance¶

Day 50-51: SageMaker Model Monitor¶

Study data quality monitoring - schema and distribution drift
Learn model quality monitoring - accuracy and performance drift
Understand bias drift detection with Clarify
Study feature attribution drift
Hands-on: Set up Model Monitor for deployed endpoint
Lab: Configure drift detection alerts
Review Notes: notes/04-monitoring-maintenance.md

Day 52-53: Observability and Retraining¶

Study CloudWatch metrics for SageMaker
Learn endpoint logging and monitoring
Understand retraining triggers and strategies
Study A/B testing for model comparison
Hands-on: Build automated retraining pipeline
Lab: Implement A/B testing with production variants

Day 54-55: Maintenance Strategies¶

Study model versioning and rollback procedures
Learn endpoint update strategies
Understand data capture for monitoring
Study cost optimization for inference
Hands-on: Implement model update with rollback capability

Day 56: Week 8 Review¶

Complete monitoring and maintenance practice questions
Design comprehensive monitoring architecture
Practice incident response scenarios

Week 9: Security, Bedrock, and GenAI¶

Day 57-58: ML Security¶

Study IAM roles for SageMaker execution
Learn VPC configuration for training and inference
Understand encryption at rest and in transit
Study compliance and data privacy for ML
Hands-on: Configure SageMaker in VPC with encryption
Lab: Implement least-privilege IAM policies
Review Notes: notes/05-security.md

Day 59-60: Amazon Bedrock and Generative AI¶

Study Bedrock foundation models and providers
Learn knowledge bases for RAG patterns
Understand fine-tuning and customization options
Study Agents for Bedrock for autonomous tasks
Study Guardrails for responsible AI
Hands-on: Build RAG application with Bedrock
Lab: Configure Guardrails for content filtering
Review Notes: notes/06-bedrock-genai.md

Day 61-62: Responsible AI¶

Study SageMaker Clarify for bias and explainability
Learn SHAP values for feature attribution
Understand model cards for documentation
Study human-in-the-loop with Ground Truth

Day 63: Week 9 Review¶

Complete security and GenAI practice questions
Review Bedrock use cases and patterns

Week 10: Final Review and Exam Preparation¶

Day 64-65: Comprehensive Practice Exams¶

Take official AWS practice exam
Complete third-party practice exams
Identify and document weak areas
Review exam feedback and explanations

Day 66-67: Weak Area Remediation¶

Review topics missed in practice exams
Re-read relevant notes and documentation
Complete additional hands-on labs for weak areas
Take targeted practice questions per domain

Day 68-69: Final Knowledge Consolidation¶

Review all notes and key concepts
Practice end-to-end ML pipeline design
Review SageMaker service limits and quotas
Take final practice exam (target 80%+)

Day 70: Pre-Exam Preparation¶

Light review of key concepts only
Confirm exam logistics and requirements
Prepare exam day materials and environment
Get adequate rest and mental preparation

Daily Study Routine (3-4 hours/day)¶

Recommended Schedule¶

60 minutes: Read study materials and AWS documentation
90 minutes: Hands-on labs with SageMaker
45 minutes: Practice questions and review
15 minutes: Note-taking and concept reinforcement

Progress Tracking¶

Weekly Milestones¶

Practice Exam Targets¶

Week 4 end: Score 60%+ on practice exams
Week 7 end: Score 70%+ on practice exams
Week 9 end: Score 80%+ on practice exams
Week 10 end: Score 85%+ consistently

Study Resources¶

Official: - MLA-C01 Official Exam Page - Registration - AWS Skill Builder - Free training and labs - SageMaker Documentation - Service documentation - SageMaker Examples - Code examples

Final Exam Checklist¶

One Week Before¶

Complete all practice exams with target scores
Review weak areas identified in practice
Confirm exam appointment and requirements

Day Before Exam¶

Light review of key concepts
Ensure technology setup works (for online proctoring)
Prepare identification and workspace
Get adequate sleep

Exam Day¶

Arrive early or log in 30 minutes before
Bring required identification documents
Stay calm and manage time (2 minutes per question)
Read questions carefully - focus on ML engineering aspects
Use elimination strategy for multiple choice