Blossom

Find clarity with Blossom, your AI mental health companion. Understand your feelings with our journal and get personal support through realistic voice conversations.

Comment

Inspiration

In today's fast-paced world, many of us struggle with stress, anxiety, and the challenge of finding a moment for genuine self-reflection. We noticed a gap in the market for a mental health tool that was not only accessible 24/7 but also deeply personal and compassionate. The inspiration for Blossom came from a simple idea: what if you could have a supportive companion to talk to anytime, one that listens without judgment and helps you cultivate your own emotional well-being? We wanted to build a tool that felt less like an app and more like a friend—a space to grow.

What it does

Blossom is a beautiful, voice-enabled mental health companion that helps users reflect, cope, and grow emotionally. It's designed to be an all-in-one tool for daily mental wellness.

  • AI-Guided Journaling: Users can write or speak their thoughts, and our AI helps them identify key themes and untangle complex feelings.
  • Sentiment Tracking: Blossom intelligently analyzes journal entries to provide users with a visual map of their emotional patterns over time, fostering greater self-awareness.
  • Compassionate Voice Conversations: This is Blossom's standout feature. Using cutting-edge AI from ElevenLabs and Tavus, users can have realistic, supportive voice conversations with their companion. It's a space to talk things through and receive empathetic guidance.

How we built it

We built Blossom as a mobile-first application with a focus on a seamless and secure user experience, leveraging a modern and scalable tech stack.

  • Frontend: The user interface was built using React Native, allowing us to create a cross-platform application with a single, beautiful codebase. We designed the UI to be calming, intuitive, and non-overwhelming.
  • Backend: A Python server (using FastAPI) handles the core logic, user data management, and communication with the various AI services.
  • AI & Machine Learning: This is the heart of Blossom.
    • We integrated the ElevenLabs API for its industry-leading, realistic, and emotive text-to-speech capabilities, giving our companion its compassionate voice.
    • The Tavus API was used to power the personalized conversational video elements, making the interaction feel more human and direct.
    • Natural Language Processing (NLP) models were used for the sentiment analysis in our journaling feature and to understand the user. Credits to Gemini AI.
  • Database: We used Firebase for its real-time database and secure authentication, allowing for rapid development and iteration.

Challenges we ran into

A hackathon is always a race against the clock, and building something as nuanced as Blossom presented several challenges:

  1. Humanizing the AI: The biggest challenge was making the AI companion feel genuinely empathetic and not just robotic. This required extensive prompt engineering and fine-tuning to ensure the tone was consistently supportive and non-judgmental.
  2. API Integration: Integrating multiple complex APIs (ElevenLabs, Tavus) under a tight deadline was technically demanding. We had to quickly learn their documentation and handle potential latency to ensure the conversations felt natural and real-time.
  3. Scope Management: Our initial vision was huge. We had to make tough decisions about which core features to prioritize to ensure we could deliver a polished, functional prototype by the deadline.

Accomplishments that we're proud of

Despite the challenges, we're incredibly proud of what we achieved in such a short time.

  • We successfully built a functional end-to-end prototype that showcases our core vision.
  • The successful integration of the ElevenLabs voice AI is our biggest technical accomplishment; hearing Blossom speak compassionately for the first time was a magical moment.
  • We created a cohesive and beautiful brand identity, from the logo and color palette to the cat mascot, which truly enhances the user experience.
  • Most importantly, we created a demo that feels gentle, safe, and genuinely helpful, proving that technology can be a powerful force for good in the mental health space.

What we learned

This project was an immense learning experience. We learned the incredible potential of generative AI to create more human-centric technology. We gained hands-on experience in rapid prototyping, integrating complex third-party APIs, and the critical importance of prompt engineering in shaping an AI's personality. Above all, this project reinforced the lesson that for sensitive applications like mental health, a thoughtful, user-focused design is just as important as the underlying technology.

What's next for Blossom

This hackathon is just the beginning for Blossom. We have a clear vision for the future:

  • Deeper Personalization: We plan to enhance the AI's memory and context awareness to create even more personalized and long-term supportive relationships.
  • Feature Expansion: We aim to introduce new features like guided meditations, personalized goal-setting based on journal entries, and breathing exercises.
  • User Feedback & Beta Testing: Our immediate next step is to get Blossom into the hands of real users to gather feedback and continuously improve the experience.
  • Pathways to Professional Help: We want to build optional pathways within the app for users to connect with human therapists and mental health professionals.

Built With

Share this project:

Updates