
Honest scope: ControlDeck NG uses mock data and is not connected to a live bank or POS system. Its AI CEO Assistant returns demo responses grounded in sample metrics. Per the repository README, the product idea, business framing and analytics are Emmanuel’s; the UI implementation was developed with AI coding assistance under his direction.
What it demonstrates
An intelligence and control layer above hospitality operations: POS, payments, inventory and expenses feed a central view, reconciliation exceptions are surfaced, and the CEO sees a dashboard plus Q&A. Exception language is deliberate: “Exception requiring management review”, never an accusation of fraud.
Screens
- CEO Home: KPIs, alerts, revenue trend, branch and unit cards.
- Financial Control Centre: payee and reference gaps, reconciliation status, exception values.
- Sales Intelligence: Restaurant, Bar, Bakery, Club/Lounge, Delivery, Campus, Other.
- Cost & Profitability: cost lines, budget vs actual, data-quality warnings.
- Inventory: stock, wastage, expected vs actual consumption.
- Purchasing: suppliers, price changes, recommended buys.
- AI CEO Assistant: natural-language Q&A over demo metrics.
- Executive Alerts: Critical / Warning / Normal.
Code structure
The source follows a page-per-screen layout: src/pages/ holds one component per screen (for example CeoHome.tsx, FinancialControl.tsx, AiAssistant.tsx), src/layout/AppShell.tsx provides navigation, src/data/mockData.ts holds the grounded mock data, and src/utils/format.ts handles naira formatting.
Grounding figures reflected in the mock data
| Metric | Value |
|---|---|
| Sales | ₦2.888B |
| Receipts | ₦2.917B |
| Payments | ₦2.116B |
| Cash/bank | ₦795.2M |
| Payments with no payee | 19,146 |
| Receipts with no reference | 33,605 |
| Food + Kitchen and Beverages | ≈ 58.9% of payments |
| Restaurant + Club/Lounge | ≈ 66.5% of revenue |
Limits
- Proof of concept with mock data; there is no live POS, bank or inventory integration.
- The AI CEO Assistant returns demo responses grounded in sample metrics; it is not a production AI system.
What this project demonstrates
- React 19 and TypeScript: an eight-screen single-page app with React Router and a shared layout.
- Data visualisation: Recharts trend and breakdown charts over typed mock data.
- Analysis-to-product thinking: screens designed around the control gaps found in the underlying analysis.
- Static deployment: Vite build published to GitHub Pages.
Run it locally
npm install npm run dev # http://localhost:5173/controldeck-ng/ npm run build
Written by Emmanuel Abolade, Data Analyst, Business Intelligence Developer and Full-Stack Software Developer based in Hull, UK. Every figure on this page comes from the linked public repository or live demo.