data-peek/docs/demo-script-notebooks.md

SQL Notebooks — Demo Script

Recording guide for the SQL Notebooks feature demo video.

Setup

• Database: localhost:5433 / acme_saas / demo / demo (Docker container datapeek-demo)
• Start container: docker start datapeek-demo
• Seed file: seeds/acme_saas_seed.sql
• Demo runbook: seeds/demo-runbook.dpnb (import this or recreate live)

Demo Flow (3-4 minutes)

Scene 1: Create a Notebook (30s)

1. Open data-peek, connect to the ACME SaaS database
2. In the sidebar, show the Notebooks section
3. Click + to create a new notebook
4. Title it "ACME SaaS Health Check"
5. Show it opens as a new tab

Scene 2: Add Markdown + SQL Cells (60s)

1. The notebook starts empty — click + Markdown cell
2. Type: # Platform Overview and a short description
3. Click away — show it renders as formatted markdown
4. Click + SQL cell
5. Type the platform overview query:

   SELECT o.plan, COUNT(DISTINCT o.id) AS orgs,
     COUNT(DISTINCT m.user_id) AS users
   FROM organizations o
   LEFT JOIN memberships m ON m.organization_id = o.id
   GROUP BY o.plan ORDER BY orgs DESC
   

6. Press Shift+Enter — results appear inline, focus moves to next position
7. Add another markdown cell: "## Revenue Health" with explanation
8. Add another SQL cell with the subscription query
9. Shift+Enter through it — show the run-and-advance flow

Scene 3: Pin Results (30s)

1. On the platform overview cell, click the ... menu
2. Click Pin result
3. Show the "Pinned — ran [date]" header appears
4. Close the notebook tab
5. Reopen from sidebar — pinned result is still there

Scene 4: Keyboard Navigation (30s)

1. Press Cmd+J / Cmd+K — show focus moving between cells
2. Press Enter on a SQL cell — enters edit mode
3. Press Escape — exits back to cell navigation
4. Press Cmd+Shift+D — deletes a cell
5. Show the bottom shortcut bar

Scene 5: Import the Full Runbook (30s)

1. Delete the notebook you just made (or create a new one)
2. Import seeds/demo-runbook.dpnb (File → Import or drag to sidebar)
3. Show the full 7-step health check runbook loads
4. Cmd+Shift+Enter (Run All) — watch the execution wave flow through each cell
5. Show results appearing one by one

Scene 6: Export & Share (30s)

1. From the toolbar ... menu, click Export as Markdown
2. Open the exported .md file — show it's readable with SQL blocks and result tables
3. Mention: "Send this to a teammate, they can read it without data-peek"
4. Show Export as .dpnb — "Or send this, and they can import it and run it themselves"

Scene 7: Duplicate to Connection (15s)

1. Right-click the notebook in the sidebar
2. Click Duplicate to connection...
3. Pick a different connection (e.g. staging)
4. Show the copy appears with "(copy)" suffix

Key Moments to Highlight

• Shift+Enter flow — the Jupyter muscle memory, run-and-advance
• Pinned results persisting — close and reopen, data is still there
• The runbook use case — "your Notion doc + SQL client, collapsed into one"
• Export as Markdown — readable anywhere, no vendor lock-in

Recording Tips

• Use dark mode (primary design target)
• Keep the sidebar open to show the Notebooks section
• Zoom to ~125% so cell content is readable in the video
• No narration needed if adding captions — keep it tight
• Target 1080p, 60fps for smooth scroll/transitions