Module Delivery Pattern — Five-Step Cheat Sheet

One-page reference for trainers

One-page reference for trainers. Every main-day module (0–7) and optional modules (8–9) follow the same five-step rhythm.

The five steps

Step	Name	Typical time	Trainer action
1	Animation	2–4 min	Play `mod-XX-*.mp4` (or narrate story beat if MP4 pending)
2	Think & Discuss	5–12 min	Use reflection-prompts.qmd for questions + trainer answers; whiteboard 3–5 bullets
3	Theory	10–20 min	Teach concept from module §3; link to trainee discussion answers
4	Quiz	2–3 min	Share Google Form (multiple choice, auto-scored)
5	Practice	Remainder	Hands-on exercises — Module 7 uses heading Practice — Open Discussion (same step, different label)

Module 7 naming: Step 5 is still practice for that module — trainee-led tool comparison replaces a coding lab. The page heading reads ## 5. Practice — Open Discussion so the five-step pattern stays consistent.

Bridge phrase (after Step 2):

“You mentioned [X]. Let’s see how the industry handles that.”

Module timing (main day)

Module	Total	Anim	Discuss	Theory	Quiz	Practice
Story Welcome	30	3	12	10	3	2
1 Fundamentals	35	3	8	12	3	9
2 Databricks	75	3	7	15	3	47
3 Snowflake	75	3	7	15	3	47
4 dbt	75	3	7	15	3	47
5 Production	45	3	7	15	3	17
6 AI Features	45	3	7	15	3	17
7 Wrap-up	30	3	5	5*	—	17†

*Module 7 theory: Objectives + Power BI notes + When to Use What only. Deep dive is self-study.

†Module 7 step 5 heading: Practice — Open Discussion (open-discussion-guide.qmd). Module 7 has no separate Quiz — the silent write + discussion replaces it.

Protect lab time: Modules 2–4 are the core — shorten discussion before cutting practice.

Environment readiness — Step 5 prep

Before releasing trainees to Practice (Step 5), the co-trainer should verify the correct portal is open and accessible for each attendee. This prevents the first 5 minutes of lab time being lost to environment issues.

All infrastructure must be provisioned before class — see pre-class-checklist.qmd and facilitator-guide.qmd §Pre-class infrastructure setup.

Module	Platform portal	Attendee should have	Co-trainer verifies
2 Databricks	Databricks Workspace	Browser tab with sidebar visible; cluster started (green dot)	Compute page — all attendee clusters show Running; both trainers ran Git → Pull on `MHP-DE-Workshop-2026` (Step 3b)
3 Snowflake	Snowsight	Own trial account created; `00_account_setup.sql` run; `DE_MASTERCLASS` database exists; `DE_WORKSHOP_ROLE` selected	Walk around — confirm each attendee completed trial signup and setup; SAS token distributed
4 dbt	Terminal (Codespaces or Docker)	`dbt_project/` directory; `profiles.yml` configured with `DE_MASTERCLASS` + `DE_WORKSHOP_ROLE`; `dbt debug --target snowflake` passing	Walk around — check terminals for green `All checks passed!`
5 Production	Databricks Jobs & Pipelines + Snowsight	Jobs & Pipelines page accessible (formerly Workflows → Delta Live Tables; renamed to Lakeflow Declarative Pipelines); Snowflake Task SQL ready	Pre-create one Lakeflow pipeline demo; `TASK_HISTORY()` returns data
6 AI Features	Snowsight + Databricks	Worksheets for Cortex SQL; Genie page accessible in Databricks (under SQL section)	Test `AI_COMPLETE` query; Genie space loads
8 Streaming	Databricks + Snowsight	Cluster with Kafka libs; Aiven topic active	Compute → Libraries tab shows Kafka jars; producer logs show events
9 ML	Databricks AI/ML → Experiments + Snowsight	AI/ML → Experiments page; `ML.FORECAST` worksheet ready	`USE AI FUNCTIONS` privilege + `CORTEX_USER` role; ML Runtime cluster available

Transition phrase (after confirming environment):

“Everyone has [portal name] open? Great — let’s start with Step 1.”

If environment check fails for multiple attendees: Extend the Theory section by 5 minutes while the co-trainer resolves issues. Never cut Practice time for environment fixes.

Priya / Power BI thread

Priya builds the dashboard in parallel with Bob’s pipeline work — not a single end demo.

After module	Dashboard beat
2 Databricks	Overview charts — trips by hour/day
3 Snowflake	Map — borough and zones
4 dbt	Revenue, payments, quality scorecard
7 Wrap-up	Full five-page walkthrough

Trainees build Gold; Priya consumes Gold in Power BI. Same schema, any engine.

Three constraints (day spine)

Introduced in Story · Revisited Module 7:

Constraint	Marcus asks
Cost	Affordable in year 3?
Performance	Fast enough for live dispatch (Module 8)?
Compliance	Audit-ready lineage by Q3?

Role boundaries (say once per day)

Elena approves tool pivots — Bob executes; trainees are Bob.
dbt runs on Snowflake — not a third warehouse.
Module 6 = Cortex LLM · Module 9 = Cortex ML (different APIs).

Optional modules (8 → 9)

Module	Story hook	Lab dataset note
8 Streaming	Live dispatch	Aiven Kafka proxy — say explicitly
9 ML	Tip prediction	Same NYC Taxi Silver

Deliver 8 before 9 when running both.

Printable timing card (cut along dashed line)

┌─────────────────────────────────────────────────────────┐
│  MODULE N — FIVE STEPS                                 │
├─────────────────────────────────────────────────────────┤
│  [ ] 1. Animation          ___ min   mod-0N-*.mp4       │
│  [ ] 2. Think & Discuss    ___ min   reflection-prompts │
│  [ ] 3. Theory             ___ min   module §3          │
│  [ ] 4. Quiz               ___ min   Google Form        │
│  [ ] 5. Practice / Discuss   ___ min   exercise OR Module 7 open discussion │
├─────────────────────────────────────────────────────────┤
│  Bridge: "You mentioned ___ — here's how we handle it." │
│  Priya beat: _________________________________________  │
└─────────────────────────────────────────────────────────┘

Document history

Date	Change
2026-06-05	Restructured from four-step to five-step delivery (added Quiz step; Think & Discuss now oral only)
2026-06-04	Added pre-class checklist reference in environment readiness section
2026-05-24	Initial four-step delivery cheat sheet