Finish and close freeCodeCamp Relational Databases
Finish what is already underway, preserve the evidence, and explicitly translate it into analyst skills.
Data Analyst route
Teaching Desk contains several old and future routes. This page is the decision layer: work only on the current stage and treat everything else as reference material.
A stage unlocks when its required tasks are done.
Finish what is already underway, preserve the evidence, and explicitly translate it into analyst skills.
Close the two packs already in progress and prove that CTEs, windows, JOINs, and aggregation work together.
Practice the actual analyst patterns: grain, cleaning, pivots, time series, funnels, retention, sessions, RFM, and attribution.
Move from exercises to a realistic dataset, validated tables, safe joins, reusable marts, and dashboard-ready exports.
Show that the analysis can be delivered to a stakeholder, not only queried by its author.
Task #1206
Task #1206
Task #1207
Task #1207
Task #1208
Task #1208
Task #1209
Task #1209
Task #1210
Task #1210
Use Python for repeatable validation, profiling, EDA, and charting after the SQL model already exists.
Task #1212
Task #1212
Task #1213
Task #1213
Task #1214
Task #1214
Task #1215
Task #1215
Task #1216
Task #1216
Task #1217
Task #1217
Task #1218
Task #1218
Turn the prepared marts into a public-safe executive dashboard with documented metric choices.
Write the business story, index the evidence, state assumptions, and make the case interview-ready.
Core PostgreSQL practice, LearnSQL, subquery repetition, Schema Design Mini, Upsert & Transactions, and PostgreSQL Modeling/EXPLAIN.
JavaScript, TypeScript, n8n, Retool, PWA, AppSheet, DevOps, algorithms, OOP, game theory, CS50, dbt, DuckDB, and Polars.
fCC Python tasks #1311-#1315 are supplements, not another mandatory course. Use them only when the CSV profiler or pandas work exposes a specific gap.
The Anti-Fraud route becomes current only for a real interview opportunity. It does not interrupt the Data Analyst route by default.
Add a compact statistics and experiment-analysis phase, then rebuild the same Olist case with DuckDB and dbt. That is the point where analytics engineering starts helping instead of distracting from the analyst baseline.