User Journal Entry (with mood analysis)
User Home Page
Gemini API recommended tips

🧠 EmotionAI: Your vibe. Your voice. Your AI.

🎯 Inspiration

EmotionAI was inspired by a simple but powerful question: “What if journaling your feelings could come with real-time emotional support — without judgment?” In a world full of stress, burnout, and constant noise — especially on college campuses — we wanted to create a calm, intelligent space where users could freely express their thoughts and receive meaningful support.

We imagined EmotionAI as a mood mirror — one that not only captures how you feel, but reflects something helpful back to you.

🛠️ How We Built It

Frontend: Built in React Native (Expo) to support cross-platform mobile journaling.

Backend: Powered by FastAPI, handling sentiment classification, AI integration, and session storage.

AI Integration: We used Google’s Gemini API to generate personalized reflections based on user input, tailored to mood and metadata (age, gender, etc.).

Sentiment Analysis: Lightweight emotional classification using TextBlob to dynamically craft the right AI tone and response.

Storage: Sessions and moods are tied to unique anonymous session_ids using AsyncStorage, with plans to scale via MongoDB Atlas.

UX Design: Simple, calming interface with color-coded emotion bubbles, AI feedback, and a journaling-first experience.

🌱 What We Learned

How to build an end-to-end AI-powered mobile app using React Native and FastAPI.
Integrating external APIs like Gemini effectively and securely.
The power of sentiment analysis in driving user experience dynamically.
The importance of frictionless UX in mental health products — every tap matters.
The challenges of balancing emotional tone with AI outputs.

🚧 Challenges We Faced

Getting Gemini API access set up correctly (v1 vs. v1beta was... a journey 😅).
Ensuring mobile-to-backend communication across devices and networks.
Designing an interface that felt personal and not clinical.
Creating flexible sentiment-based AI prompts without making the user feel judged.