How June senses urgent needs

Human attention is a terrible scheduler for crisis giving. By the time a feed ends and someone opens a donation page, another story has already taken the slot. June’s Sense stage exists so the agent can surface urgent, structured needs—and help you find relevant causes to support.

Signals, not vibes

June polls disaster and relief inputs on a schedule—including GDACS-style event feeds that describe earthquakes, floods, storms, and related alerts. Structured events give the agent something concrete to work with: an identifier, a place, a severity cue, a timestamp. That coverage is never perfect—feeds lag and miss some local crises—but it is how June stays oriented toward real-world need.

Decide: urgency with a recorded rationale

Once signals are in, Decide ranks where help may be needed now. When Gemini is configured, the model proposes an urgency score and a short rationale. When it is not, a heuristic path keeps the loop alive. The rationale explains why a need rose to the top—so chat recommendations and routing stay grounded in what June is watching.

Facts come from partners

Models narrate; charity identity and disbursement facts come from discovery partners. June uses those checks in Verify before recommending or routing support—so EINs and eligibility are not invented in prose.

From news cycle to giving

The win is reliability: you should not have to personally monitor global alerts to find where help matters. Talk with June about what you care about, or give generally so she can route to urgent verified needs. Sense and Decide power that matching; Verify and Act make it real.