Describe your project

MindCare is a cutting-edge mental health support app on Android and iOS, addressing the growing need for timely intervention in today’s world. It serves the general public and frontline healthcare workers, using advanced Machine Learning (ML) and Artificial Intelligence (AI). MindCare’s three-day assessment cycle includes analyzing questionnaires, chatbot interactions through Gemini's Natural Language Processing (NLP), biometric data from fitness trackers (Google Health APIs), and sentiment analysis from scenarios and images. This comprehensive approach enables early detection of mental health disorders.

The solution's core capabilities focus on early diagnosis and GenAI-driven remedies, like meditation, exercise, and family time, empowering users to manage their mental health. It provides mental health screenings, real-time result access, and guidance for healthcare workers. However, MindCare does not replace professional therapy or long-term psychiatric care. Its GenAI-driven suggestions are geared towards self-care and real-time guidance, but cases requiring specialized or prolonged treatment are outside its scope.

Looking ahead, MindCare holds significant growth potential. Future opportunities include expanding GenAI to provide deeper mental health insights, integrating with broader health systems, and offering virtual therapy sessions. The app could also enhance biometric tracking with Google Health APIs and refine gamification to increase user engagement, encouraging proactive mental health management.

MindCare represents a transformative tool in mental health, combining GenAI, personalized care, and a focus on security through blockchain technology for safe medical record storage.

Challenges we ran into

One of the key challenges we encountered while building MindCare was integrating the Gemini APIs for our chatbot and user report generation. Specifically, we faced issues related to:

Response Timing: The Gemini API sometimes took longer to respond, especially when handling complex mental health-related queries. This could lead to delays in chatbot conversations, affecting the user experience.
Solution: We addressed this by implementing a message queuing system to manage API requests efficiently and provide users with interim responses, such as "Processing your request," to maintain engagement during the delay.

Contextual Understanding: Since mental health queries are nuanced, the chatbot occasionally misinterpreted user inputs or provided overly generic responses.
Solution: We fine-tuned the prompt engineering to offer more specific guidance based on the user’s mental health assessment data. Additionally, we provided multiple fallback options in the chatbot flow, ensuring more accurate replies when ambiguity arose.

Report Generation Accuracy: Using Gemini to generate personalized mental health reports led to inconsistent formatting and output structure, which made it difficult to maintain uniformity.
Solution: We predefined a set of report templates and structured the data input in a more rigid format. This ensured that the generated reports followed a consistent format while still being tailored to individual users' data.

By overcoming these challenges, we were able to ensure a smoother, more reliable interaction between users and the MindCare platform.