How the Launch path is produced: a fixed weekly template, the complete ~49-week skeleton, and the assessment model. Every week is one card with eight fields; the full detail lives in the four phase pages.
Every week fills these eight fields. The structure enforces the spine: nothing is “learned” until it is built (lab) and defended (deliverable + live session).
Week number, title, module(s), phase/depth tier, and the gap(s) it closes.
3–5 measurable “by end of week you can…” statements. The bar for done.
Self-paced topics with rough hours — scaffolding for the lab, not the point.
The build task on real data: steps, dataset links, references, a starter scaffold.
What they submit (code + a “why” note) and the criteria that make it a pass.
Recap point, the likely misconception, and the “defend your decision” prompt.
Whether the week feeds the phase artifact or the capstone.
What it assumes and what later weeks build on it.
Every week mapped to its module and title. Phase bands mark depth; portfolio rows mark where a shippable artifact is assembled. Open any phase page for the full weekly detail.
| Wk | Module | Title |
|---|---|---|
| Phase 1 · Foundations · bronze · Weeks 1–13 | ||
| 1 | M00 | Dev environment, CLI & Git |
| 2 | M01 | The DE mental model & the data lifecycle |
| 3 | M02 | Python I |
| 4 | M02 | Python II |
| 5 | M02 | Python III |
| 6 | M03 | DataFrames: select, clean, transform, group & merge |
| 7 | M03 | Python for data II |
| 8 | M04 | Typed, tested, packaged: turning a script into a data product component |
| 9 | M05 | SQL I |
| 10 | M05 | SQL II |
| 11 | M06 | Advanced SQL I |
| 12 | M06 | Advanced SQL II |
| 13 | M07 | Relational theory & normalization: keys, FDs, 1NF→3NF |
| ◆ Phase 1 portfolio → a SQL + Python analytics project on real data, in a clean, documented repo. | ||
| Phase 2 · Core data engineering · silver · Weeks 14–30 | ||
| 14 | M08 | Dimensional Modeling I |
| 15 | M08 | Grain, conformed dimensions & the bus matrix |
| 16 | M09 | Recoverable History: SCD Type 2 & Effective Dating on the Provider and Price Dimensions |
| 17 | M10 | Lakehouse & gold-layer modeling |
| 18 | M11 | MDM I |
| 19 | M11 | MDM II |
| 20 | M13 | ETL vs ELT, Batch vs Incremental |
| 21 | M14 | Landing the Mess: NPPES |
| 22 | M15 | Medallion architecture |
| 23 | M16 | CDC & Incremental I |
| 24 | M16 | CDC & incremental II |
| 25 | M17 | Spark I |
| 26 | M17 | Spark II |
| 27 | M18 | Spark internals |
| 28 | M24 | Data quality I |
| 29 | M24 | Data Quality II |
| 30 | M25 | Testing taxonomy for data |
| ◆ Phase 2 portfolio → a batch medallion pipeline (bronze→gold) with SCD2 history, MDM, and DQ checks. | ||
| Phase 3 · Platform & production · gold · Weeks 31–43 | ||
| 31 | M20 | Delta Lake internals & table optimization |
| 32 | M21 | Databricks & Unity Catalog |
| 33 | M22 | Declarative pipelines I |
| 34 | M22 | Declarative pipelines II |
| 35 | M23 | Structured Streaming I |
| 36 | M23 | Structured Streaming II |
| 37 | M26 | Observability, SLAs, Lineage & Alerting: Make the Pricing-Integrity Platform Break Safely |
| 38 | M27 | Git deep |
| 39 | M28 | CI/CD for data |
| 40 | M29 | Databricks Asset Bundles |
| 41 | M30 | Governance & Security |
| 42 | M31 | Cloud foundations |
| 43 | M20 | Integration & hardening |
| ◆ Phase 3 portfolio → a production incremental + streaming pipeline with CI/CD, tests, and Unity Catalog governance. | ||
| Phase 4 · Advanced & launch · job-ready · Weeks 44–50 | ||
| 44 | M32 | Architecture decision-making |
| 45 | M12 | Elective deep-dive |
| 46 | M36 | Interview prep I |
| 47 | M36 | Interview prep II |
| 48 | M37 | Capstone |
| ◆ Capstone → the capstone data product + a recorded walkthrough you can defend in interviews. | ||
Every deliverable is graded on the same four-part rubric; each phase targets a shippable artifact — a checkpoint, not a turnstile.
The “reasoning” and “defense” lines are where the concept-vs-application gap (G1) is enforced.
Progression stays open — learners advance and catch up as time allows; the artifact is the proof, not a gate.