HealthPath

Empowering diabetes patients with personalized tools like a 3D assistant, glucose tracking, and proactive health insights for seamless diabetes care.

Comment 3

🌟 HealthPath 🌟

HealthPath is an innovative web application designed to empower diabetes patients by simplifying their health management journey. With features like a 3D assistant, personalized chatbot, glucose level tracking, and much more, HealthPath offers a comprehensive solution for diabetes care.

Sample Image

🗂️ Table of Contents

💡 Features

1. 3D Avatar Assistant

The 3D Avatar Assistant provides real-time support for diabetes management using cutting-edge AI technologies for an engaging and intuitive experience.

Key Features and How They Work:

  1. Accurate Question-Answering:

    • Uses Azure OpenAI GPT-4 and Azure AI Search to retrieve diabetes-specific information, optimized with Azure OpenAI Text Embedding for precise responses.

  2. Natural Communication:

    • Converts speech to text via Azure Speech Service, processes queries, and delivers human-like responses through Text-to-Speech, synchronized with a 3D avatar.

  3. Interactive 3D Experience:

    • A lifelike Azure Avatar enhances engagement with realistic expressions and gestures, making interactions seamless and user-friendly.

Image Description

2. Personalized Chatbot

The Personalized Chatbot in HealthPath leverages advanced AI to deliver tailored diabetes management support, making everyday decision-making easier for users.

Key Features and How They Work:

  1. Meal Analysis and Feedback:

    • Users upload meal photos, and the chatbot analyzes them to determine safety for their condition.
    • Powered by Text-Embedding-3-Large for semantic understanding of meal data.

  2. Insulin Dosage Calculation:

    • Provides dosage suggestions using the formula: CHO Insulin Dose = (Total grams of CHO in meal) / (Grams of CHO disposed by 1 unit of insulin)
    • Ensures accurate and actionable recommendations.

  3. Contextual and Structured Interactions:

    • Uses LangChain’s memory for retaining context and LangChain’s schema for structured, personalized conversations.
    • GPT-4o ensures responses are accurate and conversational.

Image Description

3. Glucose Level Tracking

  • Tracks glucose levels in real time using Continuous Glucose Monitoring (CGM) integration.
  • Displays glucose trends by day or hour for better visualization.

Image Description

4. Glucose Management Report

  • Generates a detailed report with:
    • Overview of glucose control.
    • Key patterns and trends.
    • Recommendations for improved management.

Image Description

5. Appointment Scheduling

  • Includes a calendar feature to book and manage appointments for diabetes care.

Image Description

6. Diabetes Care Blogs

  • Provides curated articles on the latest developments in diabetes management.

Image Description

7. Proactive Alerts

  • Notifies users about glucose trends and recommended actions to stay ahead of potential issues.

🛠️ Architecture Overview

HealthPath Architecture

The HealthPath platform leverages a sophisticated architecture built on Azure services to ensure scalability, robustness, and efficiency. The system is designed to streamline document processing, user interactions, and real-time data analysis.

Summary of the Flow

  1. Document Ingestion:

    • User-uploaded documents are stored in Azure Blob Storage and processed by Azure AI Document Intelligence to extract text, images, and tables.

  2. Data Embedding and Indexing:

    • Extracted data is embedded using Azure OpenAI (text-embedding-3-large) and indexed in Azure AI Search for semantic search and quick data retrieval.

  3. Performance Optimization:

    • Azure Cache for Redis enhances performance by caching frequently accessed data, while Azure AI Studio optimizes document chunking and prompt configurations.

  4. User Queries and Responses:

    • User inputs are processed through the Flask backend and integrated with Azure AI Search, CosmosDB, and LangChain for personalized and context-aware answers.
    • Responses are enriched with Text-to-Speech (TTS) via Azure OpenAI to provide a seamless conversational experience through a lifelike 3D Avatar.

  5. Monitoring and Deployment:

    • Real-time logs and performance are monitored with Azure Monitor and Application Insights.
    • The platform is hosted on Azure App Service for global accessibility and secure communication via the Application Gateway.

Image Description

🔧 How GitHub Copilot Enhanced Development

GitHub Copilot and VS Code were invaluable tools during the development of HealthPath. Here's how they helped accelerate our build process:

  1. Real-Time Code Suggestions:
    Copilot analyzed the context as we typed and provided accurate code completions. This helped us write code faster and with fewer errors, significantly reducing manual coding efforts.

  2. Multi-File Editing with Copilot Edits:
    Coordinated changes across multiple files were seamless, enabling efficient refactoring and feature implementation. This streamlined our workflow and ensured code consistency throughout the project.

  3. Natural Language Commands:
    By describing functionality in plain language, Copilot generated the corresponding code, effectively bridging the gap between our ideas and their implementation. This allowed us to focus on overall design and logic rather than low-level coding details.

  4. AI-Powered Code Explanations:
    Copilot provided explanations for complex code snippets, aiding in understanding the codebase and debugging. This feature was especially useful for onboarding team members and dealing with intricate logic.

  5. Automated Test Generation:
    Copilot assisted in creating unit tests, ensuring robust code and saving time compared to manual test writing. This enhanced the reliability and maintainability of the application.

  6. Debugging Assistance:
    Copilot identified potential bugs and suggested optimized fixes. This reduced debugging time and ensured high-quality code output.

  7. Enhanced Productivity with VS Code:
    The integration of Copilot with VS Code provided a smooth development experience. Features like Live Share enabled real-time collaboration, while extensions streamlined dependency management and deployment tasks.

By leveraging these capabilities, we accelerated the development process, improved code quality, and established an efficient workflow, allowing us to focus on building a high-impact, user-friendly application.

👥 Team

Image Description

🛠️ Challenges We Ran Into

Creating HealthPath Plus was no easy feat, and our team faced numerous challenges that tested both our technical skills and emotional resilience:

  1. Personal Stakes: This project was deeply personal for us, with one team member’s father’s experience with diabetes driving our commitment. The emotional weight of creating something meaningful added an extra layer of pressure but also fueled our determination.
  2. Complex Integrations: Merging CGM data, Azure services, and MongoDB into one cohesive platform required careful planning and collaboration. Each component needed to work seamlessly, and achieving that harmony was a true test of our engineering capabilities.
  3. Balancing Features with Simplicity: We wanted to create a powerful tool that was also easy to use. Striking the right balance between advanced features and user-friendly design was a constant challenge we worked hard to overcome.

🏆 Accomplishments That We're Proud Of

Looking back, we’re proud of what we’ve achieved as a team. Here are some milestones that stand out:

  • Turning Inspiration into Action: We transformed real-life challenges into a practical, empathetic solution that blends technology with compassion.
  • Innovative Features: Developing the 3D Assistant and Personalized Chatbot was a highlight, showcasing how AI can address real-world health needs.
  • Overcoming Hurdles Together: From debugging complex integrations to refining user flows, we tackled every obstacle with teamwork and perseverance.

🎓 What We Learned

This journey taught us invaluable lessons that extend beyond technology:

  • Empathy-Driven Design: Building with the end user in mind—understanding their struggles and needs—led to a more meaningful product.
  • Navigating Complex Systems: Integrating multiple cutting-edge tools and services deepened our technical expertise and problem-solving skills.
  • The Strength of Resilience: Every challenge strengthened our resolve and made the final product that much more rewarding.

🌍 What's Next for HealthPath Plus

We’re excited about the future and have big plans for HealthPath Plus:

  • AI-Powered Predictive Analytics: Leveraging machine learning to anticipate glucose trends and provide preventative care recommendations.
  • Dedicated Mobile App: Expanding the platform’s accessibility to ensure users can manage their health anywhere, anytime.
  • Community Building: Creating a supportive space for users to connect, share experiences, and learn from one another.
  • Global Expansion: Adapting the platform for diverse languages, cultures, and CGM devices to reach and empower more people worldwide.

Built With

Share this project:

Updates