Inspiration
Thalassemia patients face lifelong challenges β from regular transfusions and strict medication schedules to emotional stress and limited access to real-time medical support. We were inspired by real patient stories and caregivers who struggle to manage everything manually.
We asked ourselves:
What if AI could become a 24x7 companion, not just a chatbot β but a true digital caregiver?
Thus, HemoWise was born β a smart, empathetic AI health assistant built just for thalassemia warriors.
What it does
HemoWise is an AI-powered mobile app designed to support thalassemia patients in their daily health journey.
It offers:
- π€ Multilingual chatbot support (English, Telugu, Hindi) for symptoms, diet, emotional support, and FAQs
- β° Medication & transfusion reminders based on personalized schedules
- π¨ Emergency mode that detects critical symptoms and guides the patient
- π¨ββοΈ Doctor integration with real-time summary dashboards
- π² Caregiver mode for family notifications
- π Dual UI modes: Light and Dark for accessibility
- π Mood tracking to monitor emotional well-being
Everything is built to feel alive, interactive, and supportive like a digital friend that truly understands the patient's needs.
How we built it
| Layer | Tools / Tech Stack |
|---|---|
| Frontend | Flutter / React Native |
| Chatbot | GPT-3.5 API + Dialogflow (for NLP) |
| Backend | Firebase Functions + Django |
| Database | Firestore (Cloud NoSQL) |
| Auth | Firebase Auth (with role-based access) |
| Notifications | Firebase Messaging, Google Calendar API |
| Video Calls | Jitsi Meet / Twilio Video API |
We designed a chatbot that feels natural, supportive, and safe with fallback logic to connect with real doctors when needed. The emergency logic uses AI-driven symptom scoring and frequency tracking to escalate when required.
Challenges we ran into
- Ensuring AI responses remained medically safe and accurate
- Balancing empathy and clarity in multiple languages
- Designing a non-intrusive emergency button that doesnβt scare users
- Creating a multi-role system (patient, doctor, caregiver) with clean access control
- Supporting light/dark mode UI with consistent UX and visibility
- Making AI feel interactive and human not cold or robotic
Accomplishments that we're proud of
- Built a fully functional AI chatbot that understands thalassemia-specific terms in three languages
- Developed a smart symptom detection system that can classify emergency signals
- Integrated doctor dashboards that summarize real-time patient data
- Designed a clean, multimode UI with both dark and light themes
- Created a system that truly feels like a companion, not just a tool
What we learned
- How to combine AI with real human needs in sensitive domains like healthcare
- Deep insights into designing for accessibility and emotional support
- The complexity of balancing automation with medical safety
- The importance of local language support in health tech
- How to design AI that cares, not just answers
What's next for HemoWise
- π² Publish the app on Android and iOS stores
- π€ Partner with doctors and NGOs for real-world testing
- π§ͺ Integrate with EMR systems and national blood banks (like e-RaktKosh)
- 𧬠Add AI-powered insights for ferritin trends and chelation monitoring
- π£οΈ Expand to more languages and possibly voice-based interaction
- ποΈ Launch a Caregiver Companion App to support families of younger patients
Built With
- dart
- dialogflow
- django
- firebase
- firebase-functions
- firebase-messaging
- firestore
- flutter
- javascript
- jitsi-meet-api
- oauth-2.0
- openai-gpt-3.5-api
- python
- react-native
- scikit-learn
- tensorflow
- twilio
Log in or sign up for Devpost to join the conversation.