Mood Flow

Mood Flow is a personalized activity recommendation platform that suggests activities based on your mood, weather, and preferences. It also enables users to log and share activities with friends, fostering a social and engaging experience.

Inspiration

We wanted to build a project that helps people identify the perfect activity tailored to their mood and weather conditions. Additionally, Mood Flow allows users to share their consistent activities through a logging page and connect with friends.

Features

• Personalized Recommendations: Suggests activities based on user mood, interests, location, and age.
• Weather-Based Suggestions: Integrates real-time weather data for optimized activity recommendations.
• Social Features: Enables users to meet friends at suggested locations and post activities later.
• User Authentication: Allows users to log in and track their past activities.

How It Works

1. Users log in and input their mood, location, age, and interests.
2. An LLM (MistralAI 7B instruct v0.3) generates personalized activity suggestions.
3. Google Maps API provides location-based recommendations where users can meet.
4. Users can log their activities (and share them with friends: future update)

Technologies Used

• Frontend: React.js
• Backend: Express.js
• Database: MongoDB
• AI Model: Mistral-7B-Instruct-v0.3 from HuggingFace APIs for activity generation
• APIs: Google Maps API, Weather API

Challenges We Faced

• Syncing latitude and longitude data with the Weather API.
• Integrating the backend with the frontend seamlessly.
• Migration from Local LLM (due to hackathon guidelines) Ollama-3.2-3B -> Mistral AI (7B instruct) and output parsing.

Accomplishments

• Successfully built and deployed a functional application within the given timeframe.
• Explored and integrated various technologies, enhancing our technical skills.

What We Learned

• React Hooks for managing state and component lifecycles.
• Setting up and configuring the Ollama-3B model for local AI processing.
• Using Huggingface API for LLM calls.
• Implementing Express.js for handling backend API calls.
• Using GoogleMaps API.

Future Plans

• Enhanced Social Features: Expanding the networking aspect by allowing users to post, react, and share activities with each other.
• Improved UI/UX: Enhancing the interface for a more seamless user experience.
• Scalability: Improving performance and expanding the recommendation model.

Watch a tutorial and try It Out

https://www.youtube.com/watch?v=jcUoEGAPVgo