## What I Built
I built FetchAI, an intelligent, cyberpunk-styled dark mode web application that acts as a quantum dog breed scanner and an AI story generator.
When a user uploads a portrait of their dog, the application connects directly to the Google AI matrix using the latest Gemini API (gemini-2.0-flash
). It instantly extracts the exact breed classification, creates a behavioral profile, provides care metrics, and writes a beautiful, unique 3-sentence short story featuring the dog as a heroic character!
Best use of Google AI 🤖 (Powered by Gemini API via Google AI Studio).
Here is the repository link containing the full source code and preview details:
[ https://edirisinghekinurimahoshadhi-debug.github.io/Dog/] I am a 14-year-old student (Grade 9) from Sri Lanka, currently learning coding through freeCodeCamp and AI concepts via the Elements of AI course.
When I saw the "Dog Days Edition" challenge, I wanted to build something highly functional but with a premium, engaging UI. I designed the front-end using clean HTML5 and modern CSS3 featuring a glowing golden dog-paw pulse animation (🐾
) inspired by the classic Google search Easter egg.
The hardest part was converting the uploaded image into a Base64 stream inside vanilla JavaScript and passing it down to the Gemini API securely without using any heavy back-end frameworks. With strict prompts, the AI now successfully identifies the breed and automatically generates a creative narrative ("The Hero's Tale") giving the dog a random name!
This project taught me how seamlessly Google AI can be integrated into everyday client-side web applications to build fast, intelligent, and user-centric features.