Hey Product Hunt - I'm Zachary!

The raw ingredients for wildfire intelligence already exist. NASA publishes satellite thermal data every hour. NOAA streams weather and imagery. Public datasets map fuels, terrain, structures, and population exposure. But very little is connecting all of that continuously.

There’s no system that looks at a cluster of heat signatures at 4 AM and asks: is this agricultural burning or is this about to become a problem? What’s the wind doing? What’s downhill? Who lives there?

A human analyst can do that, but not continuously. Not for an entire country, especially not at 4 AM.

So I built Signet.

Signet is an autonomous wildfire intelligence agent for the continental US. It monitors wildfire activity, tracks incidents, generates outlooks, and makes predictions about fire behavior, then checks them against subsequent satellite passes.

When it sees deteriorating weather over high fuel loads with no active incidents, it can initiate its own investigation. The reasoning stays visible in a live feed instead of disappearing behind a black box.

Last month, the Texas Panhandle caught fire. Signet flagged the region as high-risk, tracked linked critical incidents, and ran through that loop autonomously for 20 hours. I didn’t touch it.

Satellites tell you where the heat is. Signet tries to tell you what it means.

It’s live now at https://signet.watch. You can sign up by ZIP code for free alerts when notable activity is detected near you.

If you check it out, I’d especially love feedback on:
- whether the nearby alerts feel useful
- whether the live situation view is clear
- whether the reasoning feed builds trust or just adds noise

And if you live in fire country, or work in emergency management, insurance, or public safety - I'd love to hear what would make this genuinely useful.

Hey Product Hunt - I'm Zachary!

Signet

Autonomous AI wildfire agent with nearby alerts

Autonomous AI wildfire agent with nearby alerts