Lumi | Devpost

Inspiration:

Mental health challenges are increasingly common in today’s fast-paced world. Many people need someone to talk to but may not have immediate access to support. We wanted to create a solution that is always available, empathetic, and non-judgmental.

What It Does:

Lumi is an AI chatbot with a friendly avatar. Users can chat with Lumi to receive comfort, reflective questions, and practical coping strategies such as mindfulness exercises or journaling prompts. Lumi helps users explore their thoughts and manage stress in a gentle, supportive way.

How We Built It:

Frontend: Lynx.js for a lightweight, interactive chat interface with the Lumi avatar. Backend: Google Gemini AI generates empathetic, context-aware responses. Security: API keys are hidden via environment variables to ensure safe AI integration.

Challenges We Faced:

Learning Lynx.js for the first time and integrating a dynamic chat UI. Ensuring AI responses were natural, empathetic, and consistent without giving medical advice.

Accomplishments:

Built a fully functional AI mental wellness chatbot with a friendly, interactive UI. Implemented reflective questions and coping strategies that are practical and safe. Successfully integrated AI streaming responses for a natural conversational experience.

What We Learned:

How to combine modern AI (Google Gemini) with a reactive frontend (Lynx.js). The importance of UX and empathy when designing mental health tools. Prompt engineering is key to keeping AI responses supportive and calm.

Next Steps for Lumi:

Improved UI: More expressive avatars and animations. New Modes: Journaling, meditation guidance, and stress relief exercises. Personalization: Tailored conversations based on user preferences for a more personal companion.