CrisisLens: AI-Powered Emergency Intelligence Radar

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the Machine.

CrisisLens is a Personal Emergency Intelligence Radar designed to help people understand emergencies through scattered and sometimes conflicting information.

During an emergency, important information can be spread across news reports, official advisories and community updates. CrisisLens brings these signals together and transforms them into actionable intelligence.

The idea behind CrisisLens is simple: people don't always need more information; they need better information.

https://github.com/divyamchoudhary1604/crisislens

CrisisLens combines open-weight AI with a modern full-stack architecture.

The frontend uses React 19 and Vite, while the backend is built using Node.js and Express.

The AI component uses Gemma 2 (google/gemma-2-9b-it), an open-weight language model. It can be run through providers such as OpenRouter, Groq or locally through Ollama.

These components work together to transform scattered emergency information into structured insights.

Open innovation gives developers the freedom to experiment, adapt and improve AI systems without depending entirely on closed proprietary models.

Using an open-weight model like Gemma 2 provides flexibility in choosing inference providers and exploring local deployment.

For an emergency intelligence platform, this flexibility matters because the system should be adaptable, transparent in its information handling and capable of evolving as new requirements emerge.

I used AI-assisted development while building CrisisLens. Add a DevRelay agent session link here if you have one.

During an emergency, uncertainty can be more frightening than the incident itself.

One report says a road is open. Another says it is blocked. Someone shares a warning, but nobody knows whether it is reliable.

CrisisLens explores the idea of an invisible intelligence working behind the scenes, connecting scattered signals and helping people understand what might be happening.

Like a ghost in the machine, it works quietly in the background, turning confusion into clarity.