MoodMentor AI

Inspiration Misinformation fatigue led us to mental health—college students like Manthan Patil, Yash Mhatre, Sahil Khan, and Harshal Deorukhakar from Pune needed a quick way to process emotions during intense hackathon prep. Inspired by journaling apps but frustrated by manual mood tracking, we built AI that instantly analyzes text and gives personalized advice. No cameras, pure Python—hackathon-built in hours!

What it does MoodMentor AI transforms journal entries/social posts into instant mood analysis (POSITIVE/NEGATIVE/NEUTRAL/EXCITED) with 90%+ confidence scores. Users get tailored advice like "Take a 5-min walk" or "Call a friend," plus CSV mood logs and n8n webhook alerts for low moods. Purple-themed Streamlit app deploys instantly on Replit.

How we built it Hugging Face's twitter-roberta-base-sentiment model powers analysis—no custom training needed. Streamlit handles the sleek purple UI with progress bars, emojis (😊😔😐✨), and Plotly mood history charts. Added rule-based advice generator + session_state for log tracking. Deployed on Replit with one-click sharing for judges.

Challenges we ran into Hugging Face model loading took 15s on Replit's free tier—added loading spinner. Short texts sometimes hit neutral zone; expanded advice rules for edge cases. n8n webhook JSON formatting needed testing across team devices in different timezones.

Accomplishments that we're proud of 90-98% confidence scores every time (pre-trained model magic). Created faceless demo video, thumbnail, and 4 test cases in <2 hours. First team deployment to public Replit—judges can test live! Mental health + AI = perfect hackathon combo.

What we learned Pre-trained Hugging Face models > custom sklearn training for hackathons (instant high accuracy). Streamlit session_state enables persistent features without databases. Team async collaboration via GitHub + WhatsApp works across Pune traffic schedules.

What's next for MoodMentor AI Chrome extension for social media auto-analysis, WhatsApp bot integration, mood streak gamification. Multi-language support (Hindi/Marathi), team dashboards, integration with Notion/Calendar for habit tracking. Open-source on GitHub—contributions welcome!

Built With

• logistic
• newspaper3k-(article-extraction)-automation:-n8n-webhook-integration-(alerts)-other:-tf-idf-vectorization
• nltk-(text-processing)-platforms:-replit-(hosting/deployment)-libraries:-pandas-(data-handling)
• plotly-(visualizations)
• python
• requests-+-beautifulsoup4-(url-parsing)
• scikit-learn-(ml-model)
• steamlist