
EchoEarth AI is a multi-perspective foresight tool that helps people reason with every stakeholder in the room, not just their own instinct. Most decision tools give you a single static forecast one lens, no room to push back. Real decisions, whether a city council weighing a desalination plant or a founder entering a new market, ripple outward and land differently on different people. You describe a decision in plain language, and EchoEarth AI generates three distinct, labeled futures: optimistic, contested, and adversarial, each grounded in 2-4 explicit assumptions. From there, you chat live with four AI personas, each arguing from their own stake in the outcome: Citizen (cost, safety, daily life), Government (regulation and public good), Business (market and ROI), and Scientist (evidence and systems). Responses stream in real time for a genuinely conversational feel. The core mechanic is editable assumptions. Change any assumption and the scenario's narrative regenerates instantly, letting you pressure-test "what if this weren't true?" without starting over. A Compare view lines up all three futures side by side, and a one-click export produces a branded PDF report with the decision, chosen scenario, and full persona transcript. Built with React 18, TypeScript, and Tailwind, backed by Supabase Edge Functions and GPT-4o-mini via the AI/ML API, with streaming handled through Server-Sent Events. The backend is stateless and key-safe by design. EchoEarth AI is built for policy makers, founders, students, educators, NGOs, and consultants anyone whose decisions affect people they can't fully see. It's honest about its limits: scenarios are AI-generated, not fact-checked; there's no persistent history; the persona set is fixed; and there's no evidence grounding yet. Next steps include branching futures, custom personas, evidence grounding, and shared team workspaces. Built by Team Bot Builders for the AI Factory Online Hackathon.
10 Aug 2026