Skip to content

Fepilot/analytics-hub-deck

Repository files navigation

📊 Analytics Hub Deck Generator

Turn your Power BI Copilot analytics reports into polished decks — in minutes.

Transform any Microsoft Analytics Hub report PDF into a tailored executive presentation or a detailed analyst interpretation guide, powered by AI analysis. No design skills, no copy-pasting, no generic output — just expert-level interpretation of your organization's actual data.


🎯 Supported Reports

Report What It Covers Best Output Mode
🤖 AI-in-One Dashboard M365 Copilot + Chat + Agents all-in-one Executive or Analyst
Super Usage Adoption Who are your super users and how they got there Executive or Analyst
🏆 Super User Impact Measurable business impact of super users Executive (ROI story)
💬 Chat & Agent Intelligence Copilot Chat sessions + Agent adoption depth Analyst
🐙 GitHub Copilot Impact Developer analytics, habit formation Executive or Analyst
📊 M365 Copilot Readiness License readiness and priority scoring Analyst

⚡ Quick Start

Step 1 — Set up (one time)

git clone https://github.com/Fepilot/analytics-hub-deck.git
cd analytics-hub-deck
code .

Step 2 — Upload your PDF

Drop your Power BI report PDF into the uploads/ folder.

How to export from Power BI:

  • Power BI Desktop: File → Export → Export to PDF
  • Power BI Service: File → Export → PDF

Step 3 — Open Copilot Chat

Open Copilot Chat in VS Code → switch mode to Agent → select Claude Sonnet (recommended) or GPT-4o.

Step 4 — Type your prompt

That's it. Keep it simple:

Create an executive deck for my AI-in-One report: uploads/my-report.pdf

Create an analyst guide for my Super Usage Adoption report: uploads/super-usage.pdf

Create an executive deck focused on the HR department: uploads/ai-in-one.pdf

Build a deck for my Super User Impact report for the CIO: uploads/impact.pdf

Dependencies install automatically on first run.


🎨 Two Output Modes

Executive Deck (executive)

For: C-suite, senior leaders, steering committees, QBRs Format: 15–20 slides, dark professional theme Includes:

  • Cover with top 6 KPIs
  • Executive summary (highest-impact findings first)
  • Dashboard screenshots with insight commentary per page
  • Cross-platform analysis or investment framework slide
  • 5 strategic recommendations backed by data
  • Closing with call to action

Analyst Guide (analyst)

For: Analytics teams, AI transformation leads, change management, IT managers, HR Format: 25–35 slides Includes: Everything in executive + deeper breakdowns:

  • Methodology and data source notes
  • Per-metric interpretation guide
  • Benchmark context (what's good, what needs attention)
  • Cohort or segment drill-downs
  • Change management action framework
  • Action planning template

📌 Prompt Reference

What you want Prompt
Executive deck, auto-detect report Create exec deck: uploads/my-report.pdf
Analyst guide, named report Create analyst guide for my Super Usage report: uploads/file.pdf
Executive for specific audience Create exec deck for our CISO: uploads/ai-in-one.pdf
Focused analysis Create exec deck focused on Agent adoption: uploads/ai-in-one.pdf
Named report + mode Create analyst guide for AI-in-One dashboard: uploads/file.pdf

The AI auto-detects the report type from the PDF content if you don't specify. If it can't determine it, you'll be asked.


📁 Project Structure

analytics-hub-deck/
├── README.md                    # This file
├── INSTRUCTIONS.md              # AI agent workflow and domain expertise
├── DASHBOARD_READING_RULES.md   # Data accuracy and interpretation rules
├── convert.py                   # Main orchestration script
├── requirements.txt             # Python dependencies
├── uploads/                     # ← Put your PDFs here
├── temp/                        # Working directory (auto-created)
│   ├── analysis_request.json    # Step 1 output
│   ├── insights.json            # Step 2 output (AI-generated)
│   └── page_*.png               # Extracted dashboard images
├── reports/
│   ├── ai-in-one/GUIDE.md       # AI-in-One expert guide
│   ├── super-usage-adoption/GUIDE.md
│   ├── super-user-impact/GUIDE.md
│   ├── chat-agent-intelligence/GUIDE.md
│   ├── github-copilot-impact/GUIDE.md
│   └── m365-readiness/GUIDE.md
└── lib/
    ├── extraction/pdf_extractor.py    # PDF → PNG + text
    └── rendering/builder.py           # PPTX assembler

🤖 Works With

  • GitHub Copilot (Agent mode in VS Code) — Recommended, easiest setup
  • Claude Code (CLI) — Also supported

Open this repo in VS Code and use Copilot Chat in Agent mode. The tool reads INSTRUCTIONS.md automatically.


📋 Step-by-Step Commands

# Step 1: Extract PDF pages (done automatically in agent mode)
python convert.py --prepare uploads/my-report.pdf

# Step 2: AI analyzes temp/ and writes temp/insights.json
# (This is done by Copilot/Claude — tell it: "Read INSTRUCTIONS.md and analyze")

# Step 3: Build the deck
python convert.py --build --output my-report-executive.pptx --mode executive
python convert.py --build --output my-report-analyst.pptx --mode analyst

🧠 Domain Expertise

This tool is built with deep knowledge of the Microsoft Analytics Hub reports. The interpretation guides encode:

  • Benchmark context: What adoption rates are good vs. concerning vs. exceptional
  • "So what?" framing: Every metric translated to business implication
  • Audience-specific narratives: Same data, different story for a CIO vs. an HR leader
  • Change management lens: What findings drive training, licensing, and enablement decisions
  • Common patterns: Typical adoption shapes, red flags, and bright-spot indicators

📚 Related Resources


License

MIT — free to use within your organization.

Built by the Microsoft Copilot ROI Advisory Team.

About

Turn any Microsoft Analytics Hub Power BI report into an executive deck or analyst interpretation guide using AI

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages