Data Engineering learning roadmap
Build the infrastructure. Design robust pipelines, manage massive datasets, and ensure data quality and accessibility.
What you will learn
- 0110h
Local Data Platform
Run the project locally, then build an API-to-PostgreSQL pipeline that validates inputs and reruns without duplicate data.
Docker & PostgreSQL · API ingestion · Idempotency & quality
- 0212h
Data Modeling
Design an analytics-ready star schema with declared grain, dependable keys, and a justified history strategy.
Grain & dimensional models · Keys & SCDs · Warehouse performance
- 0312h
dbt
Build layered dbt transformations with tested sources, documented lineage, and separate development and production targets.
Layered dbt models · Tests & lineage · Environment discipline
- 0412h
Workflow Orchestration
Schedule the platform as an observable Airflow DAG that retries safely, backfills correctly, and exposes failures.
DAGs & data intervals · Retries & alerts · Safe backfills
- 0514h
Cloud & Infrastructure as Code
Deploy the batch platform with least-privilege identities, separated environments, and documented storage and query costs.
Terraform & environments · IAM & secrets · Storage, warehouse & cost
- 0616h
Spark — Advanced
Use PySpark for a justified large-data workload, inspect its execution plan, and remove avoidable shuffle or skew.
PySpark · Partitioning & shuffles · Performance tuning
- 0716h
Real-time Streaming
Build a recoverable Kafka event flow and explain its ordering, replay, consumer-group, and delivery guarantees.
Topics · Consumer groups · Delivery semantics
- 0816h
Production Readiness Capstone
Ship the cumulative data platform with quality gates, lineage, observability, recovery procedures, and documentation another engineer can use.
Data observability · Failure recovery · Technical documentation