Module Prerequisites, Order & Editorial Guide

TRAINER ONLY

Module Prerequisites, Order & Editorial Guide

Rules for workshop-2026 content authors. Apply these when creating story-wrapped modules in workshop-2026-v2/modules/do not edit legacy files in workshop-2026-v1/.

Related: TRAINING_MATERIAL_MIGRATION_PLAN.md


Main-day module order (0–7)

# Module Depends on
0 Welcome & Setup
1 DE Fundamentals 0
2 Databricks Pipeline 1
3 Snowflake Pipeline 1 (story); same KPIs as 2
4 dbt Pipeline 2–3 (Bronze/Silver/Gold exist); story adds dbt on Snowflake
5 Production Patterns 2–4
6 AI Features 2–4 (Silver/Gold for demos)
7 Comparison & Wrap-up 2–4 (open discussion + Power BI payoff)

Each module: animation → reflection → theory → practice (see trainer/reflection-prompts.md).


Optional modules (8–9)

# Module Required Recommended Independent?
8 Streaming Modules 2–3 (batch Silver) Module 4 (dbt dynamic_table track) Yes — vs Module 9
9 Machine Learning Modules 2–3 (Silver enriched) Module 4 (dbt feature table); Gold for Cortex ML.FORECAST Yes — vs Module 8

Delivery order

When running both optional sessions, deliver Module 8 before Module 9:

  • Story: Marcus needs live dispatch (streaming) before tip-prediction ML.
  • Technical: No hard dependency — either order works if story is skipped.

Module 8 dataset (story vs lab)

Layer Content
Story / animation YellowLine NYC live taxi zone demand
Lab Aiven Kafka user-activity events — teaching proxy (same streaming patterns, different schema)

Trainer must say explicitly: “YellowLine NYC would stream taxi GPS. We use Aiven user-activity so every attendee gets a live Kafka topic without TLC streaming infrastructure.”

Do not describe Module 8 as Wikipedia edits — legacy workshop-2026-v1/index.qmd may still say that; the correct source of truth is this doc and docs/animation-production-scripts.md § Module 8.

Module 9 dataset

Same NYC Taxi Silver from Modules 2–3. Predict tip_amount on credit card trips only. No proxy dataset (contrast with Module 8).


Cortex: Module 6 vs Module 9

Trainees confuse these because both mention “Cortex.” State the distinction in Module 6 theory and again at the start of Module 9.

Module Product Examples Purpose
6 — AI Features Cortex LLM AI_COMPLETE, Copilot, Genie Analyst assistants — SQL, exploration
9 — ML (optional) Cortex ML ML.FORECAST, ML.ANOMALY_DETECTION Predictive models

Trainer line (Module 6):

“Module 6 is LLM assistants. Optional Module 9 uses different Cortex APIs for prediction.”

Trainer line (Module 9):

“This is not Module 6 again — we’re training models, not asking an LLM to write SQL.”


When migrating a module from workshop-2026-v1/workshop-2026-v2/modules/

  1. Copy technical steps from workshop-2026-v1/modules/XX-*.qmd — do not modify the original.
  2. Add at top: animation embed, reflection prompts (from trainer/reflection-prompts.md).
  3. Apply callouts from this doc (prerequisites, Cortex, dataset notes) in the new file only.
  4. Link exercises to workshop-2026-v1/exercises/ until workshop-2026-v2/exercises/ wrappers exist.

Editorial callouts to add in workshop-2026 copies

Legacy quarto module Add to workshop-2026 copy
00-welcome-setup.qmd Optional module prereqs; deliver 8 before 9
06-ai-features.qmd Not the same as Module 9 ML (Cortex LLM vs ML)
08-streaming-optional.qmd Required 2–3, recommended 4; Aiven not Wikipedia
09-ml-optional.qmd Required 2–3, recommended 4; not Module 6 Cortex
ex-streaming.qmd Module 4 recommended for dbt track
ex-ml.qmd Module 4 recommended; Gold for ML.FORECAST
setup/aiven-streaming-setup.qmd User Activity generator, not Wikipedia
setup/ml-setup.qmd Required 2–3, recommended 4

Document history

Date Change
2026-05-23 Initial editorial guide (content moved out of legacy quarto edits)
2026-05-23 Cross-reference migration plan v2.4 status