World Problem Agent

Inspiration

I was inspired to build this project after realizing that many people have great skills or domains but struggle to identify how they could tackle global-scale problems.

What it does

I built it as a fully frontend AI-powered tool using HTML, CSS, and JavaScript, where users input their domain and OpenAI generates 5 structured global problems with actionable solutions.

How we built it

Throughout the process, I learned how to handle structured JSON responses, manage user-provided API keys securely in a frontend-only environment, and design a simple but functional user interface.

Challenges we ran into

The biggest challenge was keeping the project entirely static for GitHub Pages while still connecting meaningfully to the OpenAI API, ensuring both usability and safety.

Accomplishments that we're proud of

To think about impact, I even conceptualized ideas using simple formulas like $\text{Impact Score} = \text{Reach} \times \text{Feasibility}$, which helped prioritize the most promising solutions.

What we learned

Ilearned how to integrate AI meaningfully into a static frontend, how to process and render structured JSON responses dynamically, and how to make a functional prototype that users can interact with without any backend.

What's next for World Problem Agent

Next, I plan to improve the user interface to display the ideas as readable cards, allow users to save and compare multiple domains, and eventually add a safe backend to store session data and enable collaborative exploration of global problem-solving ideas.

Built With

• ai-powered
• and
• api
• css
• for
• html
• javascript
• openai