Qloobot – Your Personalized Taste AI Chat Companion

"Qloobot – Your Personalized Culture Companion powered by AI and Taste Graphs."

Comment

Qloobot – Your Personalized Taste AI Chat Companion

🌟 Inspiration

The modern consumer is overwhelmed by choices — what to watch, where to eat, how to dress, or where to travel. We wanted to create a personalized AI chatbot that doesn’t just recommend what’s trending, but truly understands the user’s unique taste and cultural preferences.

Qloo’s Taste AI™ offers a rich dataset to personalize cultural recommendations, and Google’s Gemini LLM provides a natural way to chat with users. Combining these, we built Qloobot – a smart cultural companion that knows you.

🤖 What it does

Qloobot is an AI-powered chatbot that:

  • Understands your preferences in music, movies, food, fashion, travel, and more
  • Chats with you naturally via a sleek web interface
  • Uses the Qloo API to fetch tailored cultural recommendations
  • Uses Gemini AI to contextualize and explain those recommendations conversationally
  • Supports dynamic chat history, responsive UI, and a mobile-friendly layout

🧱 How we built it

  • Frontend: HTML5, CSS3, JavaScript (Vanilla) – Fully responsive chat interface
  • Backend: PHP scripts to call the Qloo API and Gemini AI securely
  • Hosting: Deployed on Hostinger Shared Hosting (LAMP stack)
  • APIs Used:

Tech Stack Summary

Layer Technology
Frontend HTML, CSS, JS
Backend PHP
AI Model Gemini (via API)
Data Source Qloo Taste Graph API
Hosting Hostinger (Shared Hosting with PHP)

🧩 Challenges we ran into

  • Gemini API Rate Limits: Unlike OpenAI, Gemini has strict usage rules that required optimization in prompt handling.
  • Qloo API Exploration: Understanding how to structure queries for diverse domains (music, travel, fashion, etc.) using entity IDs and types.
  • Session handling: Implementing persistent chat session logic on a PHP-hosted environment without real-time socket support.
  • UI Simplicity: Building an intuitive UI without React or Next.js, suitable for lightweight hosting environments like Hostinger.

🏆 Accomplishments that we're proud of

  • Successfully combined two advanced AI systems (Qloo + Gemini) in one seamless chat experience.
  • Built a fully hosted, functional chatbot without heavy frameworks — accessible even on low-bandwidth devices.
  • Created a system that users described as “insightfully personal” and “better than generic AI suggestions.”

📚 What we learned

  • How to leverage the Qloo Taste Graph API to map user tastes across multiple cultural dimensions.
  • Practical usage of Gemini’s LLM to wrap raw API data into fluent, engaging dialogue.
  • Hosting and integrating secure AI apps in resource-constrained shared environments.
  • Structuring user prompts and system instructions to guide LLM behavior effectively.

🚀 What's next for Qloobot – Your Personalized Taste AI Chat Companion

  • 🔗 User Accounts & Profiles – Save and evolve taste preferences over time
  • 📱 Mobile App Version – Flutter-based app for Android/iOS
  • 🧠 AI Taste Learning – Gemini fine-tuning to learn from user feedback dynamically
  • 🎨 Visual Recommendations – Display visual results (e.g., album covers, food images)
  • 🛍️ E-commerce Taste Graph – Integration with Amazon/Spotify/Netflix for real-time content/product matching

Qloobot is just the beginning of personalized AI cultural intelligence — we’re excited to grow this into a daily lifestyle companion.

Built With

Share this project:

Updates