Trainer Pre-Class Checklist

Unified checklist covering every platform, credential, and material

TRAINER ONLY

A single unified checklist covering every platform, credential, and material the trainer must prepare before a workshop delivery. Work through each section in order.

Related: facilitator-guide.qmd · module-delivery-pattern.qmd · reflection-prompts.qmd

---

2 Weeks Before — Infrastructure

Note

Azure resource group — ADLS2 and Databricks are pre-provisioned together under 1000_data_engineering_workshop (subscription MHP Resort Consulting Services, tenant mhpdev.onmicrosoft.com).

Azure / ADLS2 Storage (mhpdeworkshopsa)

Pre-provisioned — account already exists; do not create a new one.

Item	Value
Storage account	mhpdeworkshopsa
Resource group	1000_data_engineering_workshop
Location	Germany West Central
Subscription	MHP Resort Consulting Services (ba826c91-8e52-4e07-ac7c-538858bbc813)
Azure tenant	mhpdev.onmicrosoft.com

• Storage account accessible — open resource group in Azure Portal → click mhpdeworkshopsa
• Container nyc-taxi-data exists (create only if missing)
• TLC data downloaded from NYC TLC Trip Record Data:
  • yellow_tripdata_2024-10.parquet (Yellow Taxi, Parquet — workshop default month)
  • taxi_zone_lookup.csv (Auxiliary / zone lookup table)
• Uploaded to ADLS2 mhpdeworkshopsa / nyc-taxi-data:
  • raw/trips/yellow_tripdata_2024-10.parquet
  • raw/lookup/taxi_zone_lookup.csv (exact filename)
  • Portal Storage Browser or facilitator guide § TLC upload
• Container nyc-taxi-data — raw/trips/ Parquet + raw/lookup/taxi_zone_lookup.csv verified in Portal
• Storage account key (Databricks) — Security + networking → Access keys → rotate/copy key1 (facilitator guide)
• SAS token (Snowflake) — container nyc-taxi-data → Generate SAS:
  • Signing method: Account key
  • Permissions: Read + List only
  • Expiry: workshop date + 2 days
  • Copy SAS token query string (save immediately — shown once)
• Verify: Databricks spark.read.parquet(...) works with new key
• Verify: Snowflake LIST @...nyc_taxi_trips_stage returns files with SAS token

Databricks Workspace (mhpdeworkshop_databricks)

Pre-provisioned — workspace already exists in resource group 1000_data_engineering_workshop; do not create a new one.

Item	Value
Workspace	mhpdeworkshop_databricks (ID 3359135813781456)
Resource group	1000_data_engineering_workshop (same as ADLS2)

• Workspace confirmed — open resource group → mhpdeworkshop_databricks → Launch workspace

• Workspace URL noted

• You are workspace admin — can open Admin settings → Identity and access

• Git credentials linked — Settings → Linked accounts → Git integration (GitHub PAT)

• Trainer IDs agreed — pattern 00_{firstname} (facilitator guide); 2026: 00_juewei + 00_alisa

• Lead trainer — Git folder + 00_setup.py with their trainer ATTENDEE_ID (e.g. "00_juewei") → schemas 00_juewei_* created

• Co-trainer — Git folder + 00_setup.py with their trainer ATTENDEE_ID (e.g. "00_alisa") → schemas 00_alisa_* created

• Both: Home → Create → Git folder — URL https://github.com/jinjuewei/MHPDataEngineerWorkshop.git, name MHP-DE-Workshop-2026, sparse databricks/notebooks

• Both trainers Git → Pull on MHP-DE-Workshop-2026 (branch main) — workspace does not auto-sync with GitHub (facilitator guide Step 3b)

• After any new push to GitHub before class: Pull again on both trainer folders

• Groups created — workshop_trainer_2026 (lead + co-trainer) · workshop_trainees_2026 (all attendees)

• Entitlements on workshop_trainees_2026: Workspace access + Databricks SQL (facilitator guide Step 4)

• Both trainers shared folder to trainees — MHP-DE-Workshop-2026 → Share → workshop_trainees_2026 → Can Run (verify with test user)

• Users invited — each attendee accepted workspace invite email

• Unity Catalog mhpdeworkshop_databricks_2026 created (new for 2026; legacy mhpdeworkshop_databricks_2025 unchanged) — Catalog → Create catalog if missing

• Unity Catalog GRANTs on mhpdeworkshop_databricks_2026:

  GRANT USE CATALOG ON CATALOG mhpdeworkshop_databricks_2026 TO `workshop_trainees_2026`;
  GRANT CREATE SCHEMA ON CATALOG mhpdeworkshop_databricks_2026 TO `workshop_trainees_2026`;
  -- After 00_setup.py per attendee (lowercase schema names):
  GRANT USE SCHEMA ON SCHEMA mhpdeworkshop_databricks_2026.{attendee_id}_bronze TO `workshop_trainees_2026`;
  GRANT USE SCHEMA ON SCHEMA mhpdeworkshop_databricks_2026.{attendee_id}_silver TO `workshop_trainees_2026`;
  GRANT USE SCHEMA ON SCHEMA mhpdeworkshop_databricks_2026.{attendee_id}_gold TO `workshop_trainees_2026`;
  GRANT SELECT ON SCHEMA mhpdeworkshop_databricks_2026.{attendee_id}_gold TO `workshop_trainees_2026`;

• Cluster policy Workshop — max 1 worker, Standard_DS3_v2, 30 min auto-terminate; Can use for workshop_trainees_2026 + workshop_trainer_2026

• Clusters — pre-created de-workshop-{ATTENDEE_ID} or trainees self-create under Workshop policy (Step 7); verify test trainee can create cluster

• Attendees briefed: open trainer shared folder + clone 00_setup.py to Home (Databricks setup)

• SQL Warehouse de-workshop-wh — Pro or Serverless (Module 6 Genie + ai_query(); also if dbt uses Databricks target) — 2X-Small, auto-suspend 5 min; workshop_trainees_2026 + workshop_trainer_2026 Can use

• Module 6 AI — Partner-powered AI On (account + workspace); trainer dry-run Assistant / ai_query() / Genie — see facilitator guide § Step 10b

Databricks CLI & Secrets (Module 8)

• Databricks CLI installed and authenticated: databricks configure --token
• PAT (Personal Access Token) generated: Settings → User Settings → Access Tokens
• workshop-scope secrets scope created: databricks secrets create-scope --scope workshop-scope
• All 5 secrets added to scope: aiven-bootstrap-servers, aiven-ca-cert, aiven-client-cert, aiven-client-key, aiven-topic
• Secrets READ ACL — put-acl for workshop_trainees_2026 and workshop_trainer_2026 on workshop-scope

Snowflake (Trainer’s own account)

• Trainer’s own Snowflake trial account created at signup.snowflake.com
• Cloud provider: Azure, Region: West Europe
• 00_account_setup.sql executed — creates:
  • Database: DE_MASTERCLASS
  • Warehouse: DE_WORKSHOP_WH
  • Role: DE_WORKSHOP_ROLE
  • Personal schemas (e.g. 00_JUEWEI_BRONZE/SILVER/GOLD — uppercase from trainer ATTENDEE_ID)
• 01_storage_integration.sql reviewed — Option B (SAS token) used for trial accounts, no action needed
• 02_external_stage.sql executed — creates file formats and External Stages with SAS token
• Verify LIST @00_JUEWEI_BRONZE.nyc_taxi_trips_stage (your trainer ID) returns Parquet files
• (Optional) Install Snowflake CLI or SnowSQL for scripting

Aiven Kafka (Module 8 only)

• Aiven account created at aiven.io
• Kafka cluster mhp-de-workshop-kafka provisioned (Free plan, region azure-westeurope)
• SSL certificates downloaded: ca.pem, service.cert, service.key
• Service URI noted: kafka-xxxxx.aivencloud.com:12345
• Do NOT start User Activity generator yet — start on the morning of Module 8 (4-hour max)

---

1 Week Before — Materials & Credentials

Credentials Sheet

Prepare a printed card or digital document for each attendee containing:

• ATTENDEE_ID — unique identifier (e.g., 01_alice, 02_bob)
• Databricks workspace URL — or confirm invite email was sent
• Trainer notebook paths on credentials card — lead + co-trainer emails under Workspace → Users → MHP-DE-Workshop-2026
• ADLS2 Storage Account Key — to be distributed verbally (never commit to Git)
• SAS Token — for Snowflake External Stage (fresh, not expired)
• (Module 8) Aiven Service URI and SSL certificate contents

Printed / Digital Materials

• Architecture Decision Matrix — 1 per attendee
• Google Form URLs — QR codes or short links ready per module
• Open Discussion Guide — printed for trainer (Module 7)
• Whiteboard Prompts — printed for trainer

Repository

• GitHub repo 2026MHPDataEngineerWorkshop is up to date on main branch
• Production site live: https://mhp-data-engineer-2026.pages.dev/ (Cloudflare) and https://site-lemon-zeta.vercel.app/ (Vercel mirror) — CI deploys both on push to main
• Test: open the training website and verify all modules, setup guides, and exercises load correctly
• .env.template and dbt_project/profiles.yml.example have correct defaults (DE_MASTERCLASS, DE_WORKSHOP_ROLE)

Power BI (Module 7 demo + optional self-paced)

Trainer — build in Desktop (once, before class)

• Windows machine with Power BI Desktop installed (free)
• Dashboard built per Exercise: Power BI — all 12 kpi_* tables, five pages
• Data source: trainer Gold schema — Snowflake {trainer_id}_GOLD or Databricks {trainer_id}_gold (Snowflake connector)
• Import mode selected for main demo (DirectQuery only if Module 8 streaming demo planned)
• Local .pbix saved (YellowLine-NYC-KPIs.pbix) as offline fallback

Trainer — publish to cloud workspace (if you have Pro / Fabric)

• Signed into Desktop with work account that has access to your workshop workspace
• Home → Publish → report + semantic model land in the correct workspace (not accidentally My workspace)
• In Power BI Service: Refresh now succeeds — DE_WORKSHOP_WH started
• All 5 pages render in the browser, including Map (borough filled map + zone bubble map)
• Co-trainer added to workspace (if second facilitator demos Module 7)
• (Optional) Scheduled refresh configured for morning-of class
• Demo path chosen and tested: A Service browser or B Desktop screen share or C skip live demo
• If sharing report links to class: confirmed whether viewers need Pro (depends on Premium/Fabric capacity)

Trainee self-paced (after Module 4 — optional)

• Power BI setup guide and exercise linked on site
• Day-before email mentions: optional Windows + Desktop for Priya dashboard lab after dbt
• No Power BI license or account provisioning required from trainer

---

Day Before — Final Prep

• Send reminder to attendees: fork the repo, create Snowflake trial account; (optional) Windows users may install free Power BI Desktop for self-paced lab after Module 4 — see Power BI setup
• Verify attendee count matches cluster count in Databricks
• Verify ADLS2 SAS token has not expired
• Verify Aiven Kafka cluster is online (Module 8 only — do NOT start the activity generator yet)
• Test .env / Codespaces on one machine: dbt debug --target snowflake passes
• Charge laptop, prepare adapters/dongles for projector

---

Morning of Workshop (T-60 min)

• Test projector / second screen
• Open training site: https://mhp-data-engineer-2026.pages.dev/ (or Vercel / quarto preview port 4201)
• Databricks: log in to mhpdeworkshop_databricks → Compute — clusters Terminated (not Error)
• Databricks: both trainers Git → Pull on MHP-DE-Workshop-2026 (branch main) — latest notebooks from GitHub
• Databricks: both trainers’ Home → MHP-DE-Workshop-2026 — 5 notebooks; Share shows workshop_trainees_2026 Can Run
• Databricks: Catalog → mhpdeworkshop_databricks_2026 accessible
• Snowflake (trainer’s own account): log in → warehouse DE_WORKSHOP_WH resumes within ~10 sec
• Snowflake: LIST @00_JUEWEI_BRONZE.nyc_taxi_trips_stage (your trainer ID) returns Parquet files
• Snowflake: DE_WORKSHOP_ROLE visible in role selector
• (Module 8 only) Start Aiven User Activity generator (4 hours duration)
• (Module 8 only) Start relay consumer: python streaming/snowflake/00_relay_consumer.py
• Power BI: cloud workspace report refreshed or local .pbix open on Desktop
• Print materials distributed to seats
• QR codes / Google Form links tested
• Credentials cards ready to distribute

---

Credential Distribution Schedule

Item	When	How
ATTENDEE_ID	Module 1 (start of day)	Printed card or slide
Databricks workspace URL	Module 2	Email invite (sent 1–2 weeks before)
ADLS2 Storage Account Key	Module 2	Verbal or printed — never in Git
SAS Token	Module 3	Printed card — has expiry date
Aiven credentials	Module 8	Printed card — SSL certs + Service URI
Snowflake account	Self-service	Attendee creates own trial (before or during Module 3)
Databricks PAT	Module 8 (if needed)	Self-service via Settings → Access Tokens

---

Quick Reference — Who Manages What

Platform	Who creates account	Who runs setup scripts	Notes
Databricks	Trainer (workspace + invites)	Attendee runs 00_setup.py	Trainer pre-creates clusters and shares notebooks
Snowflake	Attendee (own trial)	Attendee runs 00_account_setup.sql	Creates DE_MASTERCLASS, DE_WORKSHOP_WH, DE_WORKSHOP_ROLE
ADLS2	Trainer (one shared account)	Attendee uses key in notebook	Trainer generates key + SAS token
dbt	Attendee (Codespaces or Docker)	Attendee configures profiles.yml	Targets Snowflake with DE_WORKSHOP_ROLE
Aiven Kafka	Trainer (one shared cluster)	Attendee adds secrets to Databricks	Trainer starts activity generator on Module 8 morning
Power BI	Trainer (Module 7 demo) · Attendee (optional self-paced)	Trainer builds in Desktop (12 KPIs, 5 pages), publishes to cloud workspace if licensed; attendee uses free Desktop after Module 4	Trainees: no Pro. Trainer: Pro/Fabric workspace optional but recommended for Service demo

---

Document History

Date	Change
2026-06-05	Power BI: trainer Desktop build + cloud workspace publish checklist (12 KPIs, five pages, Service demo path)
2026-06-04	Initial pre-class checklist covering all platforms and timelines