Global Pulse

Title: Global Pulse: Real-Time Instability Dashboard

Problem: Investors, journalists, and analysts need to quickly understand where global instability is emerging and how it might impact financial markets. Traditional news monitoring is slow and subjective. There's no easy way to quantify "how unstable is the world right now?"

Approach: I built Global Pulse — a real-time dashboard that transforms raw global event data into an actionable instability index. The system:

1. Ingests GDELT data — a massive, continuously-updated dataset of global news events with sentiment, location, and conflict indicators
2. Calculates a Global Instability Index (GII) — a composite score combining event intensity, volume, and acceleration, normalized against a rolling baseline
3. Visualizes instability on a 3D globe — red towers indicate elevated instability, blue indicates calm
4. Correlates with financial markets — shows how assets like gold, oil, and equities historically perform during instability spikes

Methodology:

• Used BigQuery to efficiently query GDELT's partitioned tables (850K+ events)
• Built hourly country-level aggregations with z-score normalization
• Created a weighted index: 55% intensity + 35% volume + 10% acceleration
• Pulled 2 years of ETF price data via yfinance to calculate shock-day sensitivity

Tools Used: Hex (notebooks, app builder, parameters), BigQuery, GDELT, pydeck (3D globe), yfinance, pandas

Built With

• google-bigquery
• hex
• pydeck
• python
• yfinance