Skip to content

An all-in-one Diabetes management and monitoring solution with a precise machine learning model. The user-friendly website offers daily tasks, latest blogs, an AI chatbot and health plan suggestions.

Notifications You must be signed in to change notification settings

karan-panda/Diabetes-Decoded-gdsc

Repository files navigation

Logo

Welcome to Diabetes Decoded, your all-in-one solution for diabetes management and monitoring. Our platform integrates precise machine learning models, user-friendly web applications, personalized healthcare plans, and an AI-powered chatbot to provide comprehensive support for diabetes management.

Diabetes.Decoded.User.Tour.-.Shepherd.Demonstration.Video.mp4
Know.About.Diabetes.Decoded.mp4

Table of Contents

Features😎
  • Diverse Machine Learning Models: We implement various machine learning algorithms such as Linear Regression, Logistic Regression, and KNN. Through rigorous testing, we've identified Neural Networks to provide the highest accuracy and have integrated them into our core algorithm.

  • User-Friendly Web Application: Our seamless web interface named "Diabetes Decoded" allows users to sign up, log in, and input their data for diabetes risk assessment.

  • Comprehensive Diabetes Checking: Users can input their demographic data along with advanced parameters to receive a personalized diabetes risk percentage based on our sophisticated machine learning algorithms.

  • Personalized Healthcare Plans: Based on the assessed diabetes risk percentage, users receive customized healthcare and exercise plans, along with the latest news on diabetes management and research.

  • AI-Powered Chatbot: Our chatbot, integrated with generative AI capabilities, enhances user interaction and provides personalized support. It offers advanced Natural Language Understanding (NLU), personalized responses, customized educational content on diabetes management, and adapts to different languages and dialects for global accessibility.

Work Flow🔄

Workflow

Tech Stack💻

Check out the technologies we used in Diabetes Decoded:

Frontend:

React Next.js Tailwind CSS Daisy UI

Backend:

Firebase Flask MongoDB

Machine Learning & Chatbot:

Google Colab TensorFlow Python Gemini

Version Control & Collaboration:

Git GitHub

UX Design:

Figma

Prototype (High Fidelity)🛠️

Check out our high-fidelity prototype here.

Chatbot🤖

We've developed a chatbot using the Hugging Face Llama2 model and trained it using data from the American Diabetes Association PDF. In the future, we aim to integrate Gemini or Gemma to enhance its capabilities.

Check out our project repository on GitHub: Diabetes Chatbot Repo

Installation and Setup⚙️

To set up and run Diabetes Decoded on your local machine, follow these steps:

  1. Clone the repository:

    git clone https://github.com/karan-panda/Diabetes-Decoded-gdsc.git
  2. Navigate to the project directory:

    cd Diabetes-Decoded-gdsc
  3. Install the required dependencies:

    npm install
  4. Run the Application:

    npm run dev

    The application should now be running on localhost with the frontend and backend services operational.

Our Team💪

About

An all-in-one Diabetes management and monitoring solution with a precise machine learning model. The user-friendly website offers daily tasks, latest blogs, an AI chatbot and health plan suggestions.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published