Infrastructure
Dashboards
Visualize metrics, monitor operations, and enable data-driven decisions. The interface between data and action.
📊 Sample Trading Dashboard
Handle (24h)
$2.4M
+12% vs yesterday
GGR (24h)
$186K
+8% vs yesterday
Active Users
12,450
-3% vs yesterday
Hold %
7.8%
+0.4pp vs yesterday
Handle by Hour
12am 12pm Now
Handle by Sport
NFL
45%
NBA
30%
MLB
15%
Other
10%
📺 Dashboard Types
Trading Dashboard
Users: Traders
- • Live liability
- • Sharp action
- • Line movement
⏱️ Real-time
Executive Dashboard
Users: Leadership
- • GGR
- • Handle
- • Customer count
⏱️ Daily
Risk Dashboard
Users: Risk team
- • VaR
- • Correlation
- • Concentration
⏱️ Hourly
Model Performance
Users: Data science
- • Accuracy
- • CLV
- • Calibration
⏱️ Daily
🧩 Dashboard Components
📊
KPI Cards
Key metrics at a glance
Ex: GGR, Handle, Win Rate
📈
Time Series
Trends over time
Ex: Daily revenue, user signups
📋
Tables
Detailed data exploration
Ex: Top players, open bets
🔍
Filters
Slice and dice data
Ex: Date range, sport, market
🚨
Alerts
Real-time notifications
Ex: Liability exceeded, fraud
🔎
Drill-down
From summary to detail
Ex: Click metric → breakdown
🔧 Dashboard Tools
Tableau
BI Platform
Enterprise, drag-drop
Looker
BI Platform
LookML, governance
Metabase
Open Source
Simple, SQL-based
Superset
Open Source
Python, extensible
Grafana
Monitoring
Time series, alerts
Streamlit
Python Apps
Rapid prototyping
✅ Best Practices
Do
- ✓ Start with user needs and questions
- ✓ Show context (trends, benchmarks)
- ✓ Enable drill-down from summary to detail
- ✓ Set up alerting for anomalies
- ✓ Document metric definitions
Don't
- ✗ Overcrowd with too many metrics
- ✗ Use misleading visualizations
- ✗ Ignore mobile/different screen sizes
- ✗ Forget about data freshness
- ✗ Skip performance optimization
✅ Key Takeaways
- • Design for specific user personas
- • Real-time for trading, daily for exec
- • KPIs → trends → details hierarchy
- • Alerting is as important as visualization
- • Self-service with guardrails
- • Performance matters at scale