Turn Power BI dashboards into polished, insight-driven presentations automatically. No design skills, no manual slide-building, no copy-pasting numbers.

This tool works with GitHub Copilot Chat (in VS Code) or Claude Code — pick whichever you already have.

Preferred AI model: Claude Opus or Sonnet. Set your model in Copilot Chat settings or Claude Code config.

1. git clone https://github.com/shailendrahegde/pbi-to-exec-deck.git
2. Open the folder in VS Code: code pbi-to-exec-deck
3. Open Copilot Chat → switch mode to Agent → (optional) select Claude Sonnet/Opus
4. Type:

Create exec deck "C:\path\to\dashboard.pdf"

That's it. Dependencies auto-install on first run.

git clone https://github.com/shailendrahegde/pbi-to-exec-deck.git
cd pbi-to-exec-deck
claude
> create exec deck "C:\path\to\dashboard.pdf"

Dependencies auto-install on first run. Optional alias: ./install-alias.ps1

• 16:9 widescreen slides with dark professional theme
• PBI page screenshots embedded (or vector charts with --vector-charts)
• Insight-driven headlines that answer "so what?" for each dashboard page
• Executive summary and action recommendations grounded in data
• Constitution validation — automated quality checks on the output

By default, the deck embeds PBI screenshots. Add --vector-charts to generate clean charts (bar, column, line, donut, KPI tiles, treemap, table, etc.) directly from the data.

When to use: screenshots are low-quality, you want resolution-independent visuals, or exact DAX data is available.

Create exec deck "C:\path\to\report.pbip" --vector-charts

Add context to steer the analysis toward a specific department, audience, or theme:

Create exec deck "C:\path\to\report.pbip" — focus on the Finance department

Create exec deck "C:\path\to\dashboard.pdf" — this is for the CISO; emphasize security and compliance metrics

Create exec deck "C:\path\to\report.pbip" --vector-charts — focus on May 2025 month-over-month changes

Create exec deck "C:\path\to\report.pbip" --vector-charts — focus on HR and Operations; frame recommendations around reducing onboarding time

Everything after the file path is treated as analyst guidance — adjust framing, tone, and priorities as needed.

For .pbip or .pbix files, the tool connects to Power BI Desktop via the Power BI MCP server and queries exact values using DAX.

Prefer .pbip over .pbix — PBIP exposes measure DAX expressions so the AI understands how each KPI is calculated.

python setup_pbi_mcp.py          # Download & register MCP server
python setup_pbi_mcp.py --check  # Verify installation

Downloads the official microsoft/powerbi-modeling-mcp server and registers it in .mcp.json. Runs locally — no data leaves your machine.

1. Open your report in Power BI Desktop
2. Restart your assistant session (Claude Code: /exit → claude; Copilot: reload window)
3. Create exec deck "C:\path\to\report.pbip"

Without MCP installed, PBIP/PBIX falls back to image-only analysis automatically.

MIT — use freely within your organization.

Found this useful? Star the repo to help others find it.