Crucible

Inspiration

When the Iran war started, we found ourselves struggling to understand how distant global events were directly affecting the global economy and our daily lives. We kept asking: Why are gas prices rising right now? How does a conflict across the world affect the cost of groceries or the availability of electronics? We realized that the global economy isn't a collection of isolated nations; it is a single, living body. A shock to one part—a conflict in the Middle East, a port strike in Asia, or a policy shift in Europe—sends ripples through the entire nervous system. Crucible was inspired by the need to map these hidden connections, bridging the gap between a terrifying news headline and its tangible impact on our everyday reality.

What it does

Crucible is an AI-powered geopolitical and economic intelligence platform that visualizes the world as an interconnected network. It monitors global hotspots and instantly translates complex crises into localized, tangible impacts. Instead of just showing you where an event is happening, Crucible tells you what it means. It calculates a live Global Tension Index, tracks market sentiment, and breaks down exactly how a specific disruption will cascade through the global body—predicting impacts on everything from supply chains and energy sectors to local consumer prices.

How we built it

We built the frontend using Next.js, React, and Tailwind CSS to create a sleek, immersive, and dark-themed dashboard. For the interactive global map, we utilized D3.js and TopoJSON to render a responsive, zoomable world view with custom pulsing animations for active hotspots. The core intelligence engine is powered by the Google Gemini 3.1 Pro API. We engineered complex prompts using Gemini's Structured Outputs (JSON schema) to dynamically analyze and generate realistic, interconnected global events, calculating their severity, market sentiment, and cascading economic impacts in real-time.

Challenges we ran into

One of the biggest hurdles was translating abstract geopolitical concepts into a structured, predictable data format. Getting the AI to consistently output valid JSON with logical, interconnected economic impacts—rather than just summarizing news—required rigorous prompt engineering and schema validation. Additionally, rendering an interactive, performant D3.js map that seamlessly integrates with React's state management proved technically demanding, especially when handling real-time pulsing animations, zoom behaviors, and click events simultaneously.

Accomplishments that we're proud of

We are incredibly proud of the "living map" interface. Successfully bridging D3.js with React to create a map that feels alive and instantly responsive to AI-generated crises was a major technical win. We're also proud of how effectively we utilized Gemini to turn unstructured "what-if" scenarios into quantifiable, actionable data points, successfully proving our thesis that AI can map the hidden nervous system of the global economy.

What we learned

We learned a tremendous amount about the intricacies of D3.js, geospatial data visualization, and React hooks. More importantly, working with the Gemini API taught us how to leverage LLMs not just as text generators, but as analytical engines capable of simulating complex, multi-variable economic models. Building Crucible deepened our own understanding of just how fragile and deeply interconnected our world truly is.

What's next for Crucible

Currently, Crucible uses AI to simulate highly realistic, real-time scenarios. The next major step is integrating live, real-world data feeds (global news APIs, financial tickers, maritime shipping logs) to ground the AI's analysis in absolute reality. We also plan to introduce personalized "Impact Profiles." Users will be able to input their specific industry, location, or investment

Built With

• d3.js
• geminiapi
• motion
• nextjs
• supabase
• tailwind