Feelin - Your Emotional Copilot

A compassionate AI-powered mental health companion that tracks emotions, detects crises, and keeps care teams connected in real time.

Comment

Inspiration

Mental health often goes unspoken, especially for students and working professionals who silently juggle stress, anxiety, and burnout. We wanted to build something more than a journal — a tool that listens, understands, and connects you with real help when you need it. Feelin was born from the desire to offer compassionate, intelligent support that’s always available and emotionally aware.

What it does

Feelin is an AI-powered emotional health companion that helps users track their moods, reflect on emotional patterns, and engage in supportive conversations through a chatbot. It detects crisis signals using sentiment analysis and notifies all doctors in real-time via alerts. Doctors can log in through their dedicated dashboard to view emotion logs, track mood trends, and even chat directly with patients for more personal support. This two-way interaction allows doctors to monitor how their patients are doing on a daily basis and provide timely, compassionate care. Feelin bridges the gap between users and healthcare professionals using smart technology and real empathy.

How we built it

Feelin is built using a React-based frontend and a Flask backend. We use Firebase Authentication for secure user management and Firestore as the real-time database to store user logs, roles, and chat data. The AI chatbot leverages OpenAI’s GPT-3.5-turbo model for emotionally intelligent responses, and TextBlob for sentiment analysis. We use environment variables for secret handling and implement email alerts for crisis detection using SMTP. Tailwind CSS powers the UI, and doctors/patients are routed to role-specific dashboards.

Challenges we ran into

While building Feelin, we encountered several challenges that pushed us to think critically and build smarter. Managing API keys and secrets securely, especially while working with GitHub’s push protection, required careful handling of environment variables and .gitignore configurations. Structuring Firestore to efficiently support both patients and doctors posed a unique challenge, as we had to ensure clean data access, scalability, and role-based permissions. Ensuring the reliability of real-time crisis alerts while maintaining user privacy also required thoughtful backend design. On the frontend, one of our key challenges was designing a user interface that feels safe and welcoming rather than clinical or overwhelming. Lastly, translating emotional data into actionable insights for doctors was a rewarding challenge, as we needed to present meaningful trends in a simple, digestible format.

Accomplishments that we're proud of

We’re incredibly proud of the tangible impact Feelin delivers through its thoughtfully designed features. One of our key achievements is the real-time crisis alert system that instantly notifies both doctors and crisis teams when a user shows signs of emotional distress. We built a seamless and responsive chatbot experience powered by GPT and sentiment detection, creating conversations that feel natural and empathetic. Our fully functional role-based system allows patients and doctors to interact in meaningful ways, with secure access to tailored dashboards. The clean and intuitive UI/UX ensures that users feel safe, supported, and understood. Bringing all of this together, we successfully deployed a scalable, multi-user full-stack application that blends AI, emotional wellness, and accessibility into one cohesive platform.

What we learned

Throughout building Feelin, we gained valuable insights into integrating AI into real-time applications with emotional intelligence — ensuring responses felt supportive, not robotic. We learned how to structure secure and scalable Firestore databases that support multiple user roles while maintaining data integrity. Handling GitHub push protection and managing sensitive environment variables taught us the importance of secure development practices. We also discovered how to balance technology with empathy in UX design to create a space that feels comforting and approachable. Most importantly, we deepened our technical understanding of integrating Firebase, Flask, OpenAI, and React into a cohesive full-stack application.

What's next for Feelin - Your Emotional Copilot

Looking ahead, Feelin aims to grow into an even more accessible and emotionally intelligent platform. We plan to introduce voice-based journaling combined with emotion detection, allowing users to express themselves more naturally. The chatbot will be expanded to include guided breathing exercises, daily affirmations, and mindfulness meditations to provide holistic support. Our analytics dashboard will offer weekly emotion reports for both users and doctors, making progress tracking more insightful. To make Feelin more inclusive, we intend to support multiple languages and eventually launch a mobile app using React Native for on-the-go access. Strengthening encryption and aligning with HIPAA-compliant standards will also be a key focus to ensure user data remains private and secure as we scale.

Built With

Share this project:

Updates