⚠️ MICROSOFT INTERNAL USE ONLY — This repository and all its contents are strictly confidential and intended for Microsoft employees only. Do not share externally.
AI Platform: What's Changed, What Matters, What's Next
A complete session package for the Azure Power Quarterly Business Review — built to showcase Microsoft's AI platform through rapid-fire slides and live demos.
A condensed, print-ready CSA preparation brief covering the full session flow. Includes:
- Executive Brief — 85-word crisp summary of the session
- Session Flow Table — 7 sections, 9 slides, 5 demo blocks, ~42 min
- Slide Structure & Demo Scripts — per-section slide content, demo step tables with "Show" and "Say" columns, and timing
The master session plan with full detail on every section, talking points, demo choreography, and the "trick" flow.
A fully functional demo app built to showcase GitHub Copilot agent mode — the "zero to app" super prompt concept. Built with:
| Layer | Tech | Details |
|---|---|---|
| Frontend | React 18 + React Router | 3 pages: Map View, Asset Table, Maintenance Schedule |
| Backend | Node.js / Express | 3 API endpoints with filtering, stats, and maintenance data |
| Map | Leaflet + React Leaflet | Interactive map with color-coded CircleMarkers for 20 Ontario utility assets |
| Theming | CSS Variables | Dark/light theme toggle via React Context |
| Data | Sample dataset | 20 Ontario utility assets — transformers, substations, power lines with health scores, GPS coordinates, maintenance schedules |
The session uses a cinematic build:
- ⚡ KICK OFF — Fire the GitHub Copilot super prompt into an empty folder. The audience watches it start planning.
- Present — While Copilot builds in the background, walk through Microsoft Foundry → Foundry IQ → Fabric IQ → Work IQ.
- ⚡ THE REVEAL — Come back to VS Code. The app is fully built. Walk through the working dashboard.
Empty folder → one prompt → full working app. Built while you weren't looking.
| # | Section | Time | What |
|---|---|---|---|
| 0 | Opening & Context Setting | 2 min | Frame the session |
| 1 | ⚡ GitHub Copilot — KICK OFF | 3 min | Fire the super prompt |
| 2 | Microsoft Foundry | 12 min | AI App & Agent Factory — 11K+ models, Agent Service |
| 3 | Foundry IQ | 10 min | Managed knowledge layer — agentic retrieval, cited responses |
| 4 | Fabric IQ | 4 min | Semantic intelligence — ontology, property graph (AND with Snowflake) |
| 5 | Work IQ | 1–2 min | Contextual intelligence — Work Data, Memory, Inference |
| 6 | ⚡ GitHub Copilot — THE REVEAL | 10 min | App walkthrough — map, table, maintenance, theme toggle |
# API Server
cd asset-dashboard/server
npm install
node index.js # → http://localhost:4000
# React Frontend (new terminal)
cd asset-dashboard/client
npm install
npm start # → http://localhost:3000| Endpoint | Description |
|---|---|
GET /api/assets |
All assets (filterable by status, type) |
GET /api/maintenance |
Maintenance schedule with overdue flagging |
GET /api/stats |
Dashboard stats — counts by status, type, avg health |
- Microsoft Foundry — Unified AI development platform (formerly Azure AI Foundry). 11,000+ models, Foundry Agent Service, consumption pricing.
- Foundry IQ — Managed knowledge layer. Multi-source knowledge bases, agentic retrieval with citations, ACL/Purview/Entra security.
- Fabric IQ — Semantic intelligence layer for Microsoft Fabric. Ontology workload, property graph, cross-domain reasoning. AND with Snowflake, never OR.
- Work IQ — Contextual intelligence behind M365 Copilot. Three pillars: Work Data, Memory, Inference.
- GitHub Copilot — AI-powered development platform. Agent mode: plans, codes, runs, debugs, iterates — from prompt to production.
- React 18 + React Router DOM
- Node.js / Express 4.18 with CORS
- Leaflet 1.9.4 + React Leaflet
- CSS Variables for dark/light theming
- HTML/CSS for print-ready session brief
"Build me a cloud-smart asset management dashboard for Azure Power, a utilities company. Use React for the frontend, Node.js/Express for the API, and include: a map view of assets, a status table with filtering, a maintenance schedule page, and a dark/light theme toggle. Use modern best practices, responsive design, and include sample data for 20 utility assets (transformers, substations, power lines). Set up the project structure, install dependencies, and run the dev server. Track the total build time from start to finish and display it as the first row in a final summary table that also lists the server URLs, file count with high-level types (API, components, pages, styles, data), endpoints, pages, asset count, and theme support."
One prompt. Empty folder. Full app.
Built for the Azure Power QBR — demonstrating the power of Microsoft's AI platform.