GOSafe (live at gosafe.happymonkey.ai) is our prototype for location-based disaster risk intelligence. It combines current events, official forecasts, and carefully bounded AI interpretation to answer a practical question: what should I watch around this place during these dates?
Product boundary: the current UI is a prototype with illustrative signals. It is not a live warning service and must not replace official emergency guidance.
The wedge
People plan trips, events, and moves with weather tabs, news tabs, and half-remembered map apps. GOSafe’s first slice is a location/date risk checker covering severe weather, flooding, wildfire/smoke awareness, and seismic/tsunami context — framed as situational awareness, not prediction theatre.
What “bounded AI” means here
- Prefer official forecasts and structured feeds where they exist
- Use models to interpret and summarise evidence, not invent certainty
- Refuse unreliable claims such as precise earthquake prediction
- Keep the product honest about prototype vs operational warning systems
How it is built
GOSafe sits in the disaster-intelligence stack as a modern Next.js app with a clear documentation spine (CONTEXT, ADRs, plans, research). Infrastructure pieces include nginx vhosts for the public hostname and process management for the live prototype.
Why we are building in public
Disaster-adjacent products attract hype and liability. Publishing the prototype boundary early is part of the product: show the architecture, show the limits, and iterate with source-backed signals rather than fear marketing.
Live prototype: gosafe.happymonkey.ai
Theme: evidence-first risk awareness for travel and everyday planning.