A data architect designs how an organization's data is stored, modeled, moved, secured, and used. You do not start here. You start by learning SQL and relational modeling, become a solid data engineer or database developer, and then grow into designing whole systems. The path has three stages: Foundations (SQL, modeling, databases), Job-ready (warehouses, pipelines, cloud, performance) and Advanced (architecture strategy, governance, streaming, and leading standards). Everything here uses free resources. Work through each project and put it on GitHub, because hiring managers want to see design
See the full roadmap free
Every stage from beginner to expert, with lessons, practice drills, projects, interview questions and free resources. Free with a Rolearch account.
Stage 1 of 3 · about 10 weeks
Write confident SQL and design a clean relational schema from a business description.
SQL querying
Every architecture decision ends up as queries. If you cannot read and write SQL fluently, you cannot judge whether a design is good.
Relational data modeling and normalization
A model is the core deliverable of your job. Normalization (1NF to 3NF) prevents duplicate and contradictory data, and you must know the rules before you can break them on purpose.
Database internals: indexes, transactions, ACID
Performance and reliability problems come from how databases store data and handle concurrent writes. Knowing B-tree indexes and isolation levels lets you predict problems before they happen.
Diagramming and documentation
Your designs only work if other people understand them. Clear diagrams and written decisions are a large part of the job.
Then 2 more stages (about 46 weeks in all) with lessons, drills, projects, 6 interview questions with strong answers and free resources.