Are Your People Actually In Its Way?
Part 3 of Inside the Risk Model: how ShadowIQ decides whether an event touches your trip at all — matching itineraries in place and time, and failing safe when it can't locate an event.
A magnitude-6 earthquake is a serious, live event. If it strikes a city your group left last week, its risk to your trip is zero.
That gap — between how serious an event is in the abstract and whether it touches your people — is what the exposure axis closes. It's Part 3 of Inside the Risk Model, after consequence and likelihood.
The generic-alert problem
Most travel alerting works at the country level: there's an event in France, you have people in France, you get an alert. But "in France" is not "in the way." A strike in Marseille is irrelevant to a group in Paris. A flood warning that clears before your team arrives is noise. Country-level matching is why so many risk feeds become background hum that people learn to ignore.
ShadowIQ scores exposure as a specific question: does this event intersect a traveller's actual itinerary — in place, and in time?
Place and time, together
Exposure combines three things:
- Place — how close is the event to where the itinerary actually goes? Same site is full weight; nearby, same city, and further out step down; genuinely unrelated is zero.
- Time — is anyone scheduled to be there while the hazard is live? An event overlapping a segment counts fully; one outside the travel window is heavily discounted.
- Location certainty — how precisely could we place the event? A pinpoint report counts more than a vague regional one.
Multiply those together and you get a relevance score. An event has to clear a meaningful bar on all three to weigh on your trip — serious-but-distant, or serious-but-after-you've-left, drops away.
Based on the itinerary, not on tracking people
Worth being precise here, because it's a common misconception: exposure is computed from the trip's planned itinerary — the segments, dates, and places you've told us about — not from tracking anyone's live GPS position. This is deliberate. It fits how duty-of-care programs actually plan (against itineraries), and it supports low-PII group travel, where you may not be holding individual location data at all. ShadowIQ tells you which trips and travellers are scheduled to be in an affected place and window — which is the question a coordinator can actually act on.
Failing safe when we can't locate an event
Here's a subtle but important design choice. Sometimes a source reports an event without a clear location — a vague dateline, an unresolvable place name. A naive system might "guess" it onto the nearest itinerary point. That's dangerous in both directions: invent a threat that isn't there, or attach a real one to the wrong place.
ShadowIQ fails safe: if we can't establish where an event actually is, we do not fabricate a match onto your route. Instead, a serious-but-poorly-located report is routed to a human analyst to place it properly. We would rather have a person resolve an ambiguous location than have the machine quietly assign zero-distance to a guess.
Why it matters to you
Exposure is the axis that turns a global event feed into your risk picture. It's why a ShadowIQ trip doesn't light up over every incident in the country — and why, when it does light up, it's because something genuinely intersects where your people are meant to be. Fewer, truer alerts is the entire point.
Next in the series: Do We Even Trust This Report? — how source reliability and corroboration route an alert without inflating its risk.