casual.click

Inspiration

Modern search engines and recommendation systems are designed to show users more of what they already know or expect. While this is useful, it can also limit discovery across different fields of knowledge. I wanted to explore the opposite idea: what happens when completely unrelated domains are intentionally brought together? That question inspired casual.click, a project designed to generate unexpected but meaningful connections through a single click.

What it does

casual.click is an AI-powered semantic collision engine. With one click, it selects two distant domains—such as climate science and artificial intelligence, architecture and boxing, or healthcare and finance—and generates a structured insight that bridges them.

Each discovery includes:

• A unique cross-domain concept
• A concise insight
• Why the idea matters
• A potential opportunity
• A direction for further exploration

The goal is not to answer a question, but to spark new thinking, creativity, and innovation.

How I built it

I built casual.click using:

• Python
• FastAPI
• Google Gemini API
• Tailwind CSS
• HTML, CSS, and JavaScript

The backend randomly selects two domains from a curated knowledge pool and sends them to Gemini with a structured prompt. Gemini generates a novel connection and returns a JSON response containing the insight, significance, opportunity, and exploration path. FastAPI serves both the API endpoint and the user interface, creating a lightweight one-click experience.

Challenges I ran into

One challenge was ensuring that AI responses remained structured and consistently usable. Early versions sometimes returned markdown-formatted JSON or responses that were too long for a clean user experience.

Another challenge was balancing simplicity and depth. The interface needed to stay true to the domain name—casual.click—while still producing meaningful and thought-provoking discoveries.

I also developed the project under significant time constraints, which required focusing on core functionality and reliability before additional features.

Accomplishments that I am proud of

• Building a complete end-to-end AI application as a solo participant.
• Creating a working semantic collision engine that produces unique discoveries on every click.
• Designing a simple and intuitive user experience that matches the project's domain name.
• Successfully combining AI, backend development, and frontend presentation into a single lightweight application.
• Transforming a random domain challenge into a meaningful exploration tool.

What I learned

Through this project, I learned that innovation often emerges at the intersection of unrelated ideas. I also gained practical experience working with FastAPI, structured AI prompting, JSON-based outputs, and rapid prototype development.

Most importantly, I learned how important prompt design and response validation are when building AI-powered applications that need consistent results.

What's next for casual.click

Future versions of casual.click could include:

• Semantic distance scoring between domains
• Visual knowledge maps of discovered connections
• User-selected discovery themes
• Saving and sharing discoveries
• Community-generated discovery collections
• Integration with external knowledge sources and research databases

My long-term vision is to evolve casual.click into a discovery platform that helps people explore ideas beyond their normal information bubbles and uncover unexpected opportunities for learning, creativity, and innovation.

Built With

• fastapi
• gemini
• git
• github
• html5
• javascript
• python
• tailwind
• uvicorn
• vscode
• windows-10