How RadarSignal360 works
Data → Normalization → Signals → Score → Explanation → Alerts. Every step is designed to be reproducible, testable, and auditable.
1.Data
We ingest fundamental data from primary and official sources -- government agencies, trade organizations, climate datasets, and shipping activity -- across supply, demand, inventories, flows, positioning, weather, macro and geopolitical risk.
2.Normalization
Raw values are converted into comparable signals: period-over-period change, year-over-year change, z-scores, historical percentiles, deviation from seasonal averages, and more. Units are never compared directly.
3.Signals
Normalized sensor values are grouped into category signals -- supply, demand, inventories, flows, positioning, weather, macro, geopolitics, shipping, and refining.
4.Score
Category signals are combined using configured weights into a single 0-100 score, with 50 as neutral. This is entirely deterministic application logic -- not AI.
5.Explanation
OpenAI is used only to turn the structured, already-calculated facts into a concise, plain-language explanation of what changed and why. It never decides the score itself.
6.Alerts
When a followed market's conditions change meaningfully, you're notified in-app and (if enabled) by email -- deduplicated so you aren't spammed about the same underlying event.