🤖 FlexiFlow: AI-Powered Personalized Support for Arthritis Management

FlexiFlow is a cross-platform web and mobile application designed to support arthritis patients with adaptive exercise recommendations, intelligent reminders, and a gamified self-care experience. Built with Flutter, FastAPI, and Firebase, it uses a contextual bandit (LinUCB) model to personalize care and evolve with the user's behavior.

---

🧠 Core Features

• 🏃‍♀️ AI-Powered Exercise Recommendations
  Starts with expert rules (EULAR) and transitions to LinUCB model for personalized plans.

• ⏰ Smart Reminders
  Medication reminders categorized as Upcoming, Due, Missed, and Overdue.

• 🏅 Gamification
  Badges, daily streaks, and progress tracking to motivate user consistency.

• 🗣️ Community Forum
  In-app, real-time social support with offensive word filtering.

• 📊 Multi-log Tracking
  Pain, medication, diet, and exercise logs synced in real-time.

• 🧬 Real-time Adaptation
  Personalized suggestions evolve based on feedback and behavior.

---

⚙️ Architecture

• Frontend: Flutter (for Web + Android)
• Backend: FastAPI (deployed via Render)
• Database: Firebase Firestore
• AI Model: LinUCB Contextual Bandit Algorithm (Python)
• Hosting: Live Web App

---

🔁 AI Personalization Workflow

1. Cold Start: Based on EULAR guidelines.
2. Dynamic Adaptation: LinUCB adjusts based on:
   • Pain level
   • Mood
   • Joint affected
   • User feedback (explicit or engagement-based)

---

📱 UI Modules

• Dashboard (progress view, logs)
• Exercise Planner (personalized plans)
• Logs (pain, diet, medicine)
• Forum (user posts, deletions, filters)
• Notification Center
• Profile with badges & streaks

---

🧪 Testing and Results

• ✅ Functional testing across platforms using Postman & Swagger
• ⚡ Fast API responses (~300–500ms)
• 🔄 Real-time sync with Firebase Firestore
• 💡 AI model effectiveness validated through user feedback loops
• 🎮 High user engagement with gamification and simple UI

---

🔮 Future Scope

• 🌐 Multilingual and offline support
• 🩻 Wearable & sensor integration
• 🤖 AI-based pain prediction & chatbot assistant
• 📈 Visualizations for caregivers & clinicians
• 💬 In-app video-based exercise guidance

---

🩺 Reviewed By

Dr. Veena (MBBS, MD)
Chanre Rheumatology & Immunology Center
“I would like to recommend this product if commercially launched.”

---

👩‍💻 Contributors

• Trisha Vijayekkumaran
• Bhavika Gandham
• Nagasarapu Sarayu Krishna

---

📜 Citation

FlexiFlow: An AI-Powered Personalized Support System for Arthritis Management
Presented for 19CSE314 – Software Engineering, Amrita Vishwa Vidyapeetham, Bengaluru

---

📄 License

This project is intended for academic, research, and demo purposes. Forks and contributions are welcome with attribution.