SympSolver is an AI-powered healthcare platform that provides real-time preliminary diagnosis, risk assessment, and personalized medical recommendations based on user-reported symptoms. The platform combines a Next.js web interface, Flask backend processing, a Support Vector Machine (SVM) model for disease prediction, and a Dialogflow-based conversational chatbot for interactive, accessible, and proactive health management.
SympSolver enables users to input their symptoms via a dropdown interface or an integrated chatbot and instantly receive:
- Probable diagnoses (SVM-based)
- Personalized health advice
- Lifestyle modification suggestions
- Emergency alerts for critical symptoms
- Nearby hospital recommendations via Google Maps integration
The system leverages a Support Vector Machine (SVM) model for highly accurate disease prediction and continuously refines its outputs based on updated user data.
- SVM-Based Symptom Analysis: Instant preliminary diagnosis using the SVM model with ~98% accuracy.
- Conversational Chatbot: Integrated using Dialogflow for symptom input, health queries, and follow-up advice.
- Real-Time Predictions: Backend Flask API processes inputs and serves live predictions.
- Emergency Alerts: Dynamically flags critical symptoms like chest pain or shortness of breath.
- Nearby Hospital Locator: Embedded Google Maps and OpenStreetMap to find nearby medical facilities.
- Downloadable PDF Reports: Users can download a summary of their diagnosis and recommendations.
- User-Friendly Web Interface: Built with Next.js and Tailwind CSS for responsive, smooth interaction.
- Dynamic Critical Symptom Flagging: Provides immediate emergency alerts when necessary.
- Next.js – React-based framework for building fast and scalable UIs.
- Tailwind CSS – Modern, utility-first CSS framework.
- Axios – For efficient API communication.
- Google Maps API & OpenStreetMap API – For hospital location services.
- Flask (Python) – REST API to process symptoms and serve ML predictions.
- Pickle – Used to serialize and load trained ML models.
- Support Vector Machine (SVM) – Selected for its superior accuracy (~98%), precision, and recall.
- Additional ML Models Evaluated: Random Forest, XGBoost, CatBoost, KNN, Multinomial Naive Bayes (for comparison).
- Dialogflow – Conversational agent for symptom collection and user interaction.
├── app/ # Next.js frontend structure
├── components/ # UI components (symptom input, chatbot, result display)
├── public/ # Static assets (images, icons)
├── styles/ # Tailwind CSS configuration and styling
├── backend/ # Flask backend serving prediction API and ML models
├── requirements.txt # Python dependencies for backend
├── README.md # Project documentation
└── package.json # Frontend dependencies
cd backend
pip install -r requirements.txt
python app.pyThe backend will be available at http://localhost:5000
npm install
npm run devThe frontend will run on http://localhost:3000
- Import the provided Dialogflow agent in your Dialogflow console.
- Connect Dialogflow API to your frontend using the webhook integration provided.
- Mobile Application Support
- Multilingual Chatbot Integration
- User Authentication & Symptom Tracking
- Integration with Electronic Health Records (EHR)
- Advanced Predictive Analytics for Long-Term Health Monitoring
- Deployment to Vercel / Netlify and Cloud Hosting for Flask API