Orchestrated a multi-source real-time data pipeline: Open-Meteo (weather + air quality, no key required) delivering live temperature, humidity, wind, precipitation, PM2.5, PM10, AQI, ozone, and NO₂. News intelligence routed through a Netlify serverless proxy to Google News RSS feeds. All APIs called directly from the browser — zero server latency, zero cold-start delays. Built with the same data architecture precision required to monitor environmental compliance across a 3,200+ node national infrastructure.

Cross-city intelligence briefings generated by Claude Haiku (Anthropic), orchestrated through a secure Netlify serverless function. When a briefing is triggered, the platform assembles live environmental data from all selected cities and routes it to Claude via API — the key stored server-side, never exposed to the client. The AI layer applies the same agentic workflow model used across every client engagement: define the intelligence objective, orchestrate the data inputs, generate the output.

Originally a Python/Streamlit application on Streamlit Community Cloud — re-architected as a production-grade static deployment on Netlify to eliminate 5+ minute cold-start delays. Plotly.js orchestrates all chart rendering including radar profiles and geo-mapped AQI visualization. Deployed via GitHub → Netlify CI/CD pipeline. Built through an agentic AI workflow — engineered for scale, not for demos.

Governance: API key isolated server-side via Netlify serverless function — zero client exposure. AI briefing output follows a structured 4-section format (conditions, concerns, patterns, takeaway), ensuring consistent, auditable intelligence output on every call.