Luna - my AI bestie

Your AI bestfriend!

Comment

Luna - My AI Bestie 🤖💫

Inspiration

We wanted to create a personal AI assistant that feels like talking to a real friend. Instead of generic recommendations, Luna gets to know you personally through your music playlists, food orders, watch list, and preferences to give you recommendations that actually match your taste.

What it does

Luna is a WhatsApp bot that becomes your personal AI bestie! She:

  • Gets to know you through a friendly onboarding process
  • Analyzes your Spotify playlists to understand your music taste
  • Learns your food preferences and location
  • Uses Qloo AI to give you personalized restaurant and music recommendations
  • Remembers your preferences and gets smarter over time
  • Talks like a real friend with warm, friendly messages

How we built it

  • WhatsApp Integration: Used Twilio's WhatsApp API to create the chat interface
  • AI Recommendations: Integrated Qloo API for intelligent food and music suggestions
  • Personalization: Built a user profile system that stores preferences and shared data
  • Data Processing: Created a system to analyze Spotify playlists, DoorDash orders, and other user data
  • Flask Backend: Python Flask server handles all the bot logic and API calls
  • User Experience: Designed a conversational flow that feels natural and friendly

Challenges we ran into

  • Qloo API Limitations: The API sometimes returned the same recommendations regardless of mood/filters
  • WhatsApp Button Integration: Twilio's interactive buttons were complex to implement dynamically
  • Data Processing: Extracting meaningful patterns from user data (playlists, food orders) was tricky
  • API Timeouts: Some complex Qloo queries took too long and timed out
  • User Flow Design: Making the onboarding process feel natural and not overwhelming

Accomplishments that we're proud of

  • Created a truly personalized AI assistant that remembers user preferences
  • Built a friendly, conversational interface that feels like talking to a friend
  • Successfully integrated multiple APIs (Twilio, Qloo) for a seamless experience
  • Implemented a smart data processing system that learns from user behavior
  • Designed an onboarding flow that makes users comfortable sharing their data
  • Made AI recommendations feel personal and relevant to each user

What we learned

  • How to build conversational AI that feels human and friendly
  • The importance of user experience in AI applications
  • How to handle API limitations and work around them
  • The value of personalization in recommendation systems
  • How to process and analyze user data to create meaningful insights
  • The challenges of integrating multiple third-party services

What's next for Luna - my AI bestie

  • Movie Recommendations: Add Netflix watch history analysis for personalized movie/TV show suggestions
  • Real-time Location: Use GPS to suggest nearby restaurants and events
  • Social Features: Let users share recommendations with friends
  • Voice Integration: Add voice messages and voice-activated commands
  • Learning Algorithm: Make Luna smarter by learning from user feedback and interactions
  • More Platforms: Expand beyond WhatsApp to other messaging platforms
  • Advanced Personalization: Add mood detection and context-aware recommendations

Built With

Share this project:

Updates