An analytics dashboard with KPI tracking, revenue trend charts, order breakdowns and product rankings β plus a chat that shows how questions in plain language can be answered from the data. All figures belong to a fictional company, and the chat answers are prepared for this demo; in a client project it is connected to an AI model and the real database.
Lumina Analytics
Fictional sample company Β· demo data Β· Analyzing: Jan β Dec 2025
$1,302,400
Annual Revenue
16,848
Total Orders
+18.4%
YoY Growth
Jan β Dec 2025 Β· Full year dataset
KPI Overview
Total Revenue
π°$1,302,400
+18.4%Total Orders
π¦16,848
+14.2%Avg Order Value
π$77.30
+3.7%Conversion Rate
π―4.1%
+0.4ppNew Customers
π₯2,840
+22.6%Churn Rate
π1.6%
-0.3ppCustomer LTV
β$620
+8.2%Repeat Purchase
π71%
+5.1%YoY Growth
π+18.4%
+2.8ppRevenue / User
π΅$458
+6.3%Revenue Trends
Revenue Trend
Revenue Over Time Β· 30D
Orders & Comparisons
Breakdown
Order Status Mix
Comparison
Dec vs Nov (weekly)
Products & Correlations
Products
Top by Revenue
Correlation Β· Last 30D
Orders vs Revenue (per day)
Transaction History
Orders
Recent Transactions
| Order | Customer | Amount | Status | Date |
|---|---|---|---|---|
| #ORD-9481 | TechVision GmbH | $1,200 | Completed | Dec 28 |
| #ORD-9480 | DataFlow AG | $399 | Processing | Dec 28 |
| #ORD-9479 | CloudMind BV | $4,800 | Completed | Dec 27 |
| #ORD-9478 | Pulse Analytics | $149 | Pending | Dec 27 |
| #ORD-9477 | Axis Digital | $1,200 | Completed | Dec 26 |
| #ORD-9476 | Nexo Systems | $299 | Refunded | Dec 26 |
| #ORD-9475 | BrightData Ltd | $399 | Completed | Dec 25 |
| #ORD-9474 | Sigma Insights | $4,800 | Completed | Dec 24 |
Full Dataset
Key Insights
AI Insights
Analysis Β· 30D
Click Analyze to see insights for the 30D period.
Ask Your Sales Data
AI Chatbot
In a client project
How the chat works when it is connected
You ask a question
Type any natural language question about your sales data β no SQL needed.
Claude decides what to fetch
The AI reads your question and picks the right database tool (sales by period, product performance, KPIs).
Tool queries the database
A server-side function runs the query against your real database and returns structured JSON.
Claude answers in plain English
The AI receives the data and writes a clear, human-readable response with numbers formatted nicely.
/api/chat (Next.js server) β Claude with 3 tools (query_sales_by_period, query_product_performance, query_kpis) β mock DB β Claude formats answer β Browser. The agentic loop means Claude can call multiple tools in one response if the question needs combined data.